NO304267B1 - Analogous Process for Preparation of a Therapeutically Active Specific Platelet Aggregation Inhibitor (PAI) Peptide - Google Patents

Abstract

An assay for screening snake venom for the presence or absence of platelet aggregation inhibitors (PAIs) based on specific receptor binding is described. Using this assay, the identification and characterization of PAIs in a wide range of snake venom samples was accomplished. The isolated and purified PAI from several of these active snake venoms is described and characterized. In addition, PAIs lacking the Arg-Gly-Asp (RGD) adhesion sequence but containing K*-(G/Sar)-D wherein K* is a modified lysyl residue of the formula R<1>2 N(CH2)4CHNHCO- wherein each R<1> is independently H, alkyl(1-6C) or at most one R<1> is R<2>-C=NR<3> wherein R<2> is H, alkyl(1-6C), phenyl or benzyl, or is NR<4>2 in which each R<4> is independently H or alkyl(1-6C) and R<3> is H, alkyl(1-6C), phenyl or benzyl, or R<2>-C=NR<3> is a radical selected from the group consisting of: <CHEM> where m is an integer of 2-3, and each R<5> is independently H or alkyl(1-6C); and wherein one or two (CH2) may be replaced by O or S provided said O or S is not adjacent to another heteroatom are prepared and shown to specifically inhibit the binding of fibrinogen or von Willebrand Factor to GP IIb-IIIa.

Description

The present invention relates to an analogue method for the production of a therapeutically active specific platelet aggregation inhibitor (PAI) peptide with the ability to inhibit the binding of Fg or vWF, GP Ilb-IIIa with significantly more, i.e. that either the percentage inhibition is at least two times higher at a given concentration of PAI or that the concentration of PAI causing 50$ inhibition is at least twofold less for Fg or vWF/GP Ilb-IIIa binding inhibition than for alternative ligand/receptor binding.

The peptides are related to platelet aggregation inhibitors, isolated and purified from various snake venoms. These peptides are useful as therapeutic agents for the treatment of, and prevention of, platelet-associated ischemic disorders. They block specific receptors for adhesive proteins involved in the adherence and aggregation of platelets.

Heart disease is the main cause of death in most western countries. Deaths caused by heart disease are often induced by platelet-dependent ischemic syndromes that are initiated by atherosclerosis and arteriosclerosis and include, but are not limited to, acute myocardial infarction, chronic unstable angina, transient ischemic attacks and strokes, peripheral vascular disease, arterial thrombosis, preeclampsia, embolism, restenosis and/or thrombosis after angioplasty, carotid endarterectomy, anastomosis of vascular grafts, and chronic cardiovascular devices (eg, indwelling catheters or lead "extracorporeal circulating devices"). These syndromes represent various stenotic and occlusive vascular disorders that are believed to be initiated by platelet activation either on vessel walls or within the lumen by blood-borne mediators, but are manifested by platelet aggregates that form thrombi that inhibit the flow of blood.

Many studies have contributed to an understanding of the mechanism of platelet aggregation and thrombus formation. Platelets react to various damages to blood vessels, such as narrowing of the opening, plaque formation and the presence of foreign bodies (e.g. catheters), etc. The platelet response in these injuries is a sequence of events that includes platelet adhesion and activation, and the release of platelet components, including potent cellular mitogenic factors. The activated platelet aggregates induce the formation of fibrin, which further stabilizes the thrombi.

Much is now known about the mechanisms that regulate these responses. Although unstimulated platelets contain receptors for several adhesive proteins including laminin (VLA 2, VLA 6) and collagen (VLA 2, GPIV, others), the initial coupling of platelets to the subendothelium is thought to be mediated by the binding of the platelet membrane glycoprotein (GP) Ib to the immobilized von Willebrand factor. Subsequent activation of platelets can be initiated by one or more known physiological agonists including: ADP, ephenifrine, thrombin, collagen and thromboxane A2.

Platelet aggregation is mediated by the GP Ilb-IIIa complex on the surface of the platelet membrane. GP Ilb-IIIa exists on the surface of unstimulated platelets in an inactive form. When platelets are activated by adhesion and the physiological agonists, GP Ilb-IIIb is also activated to become a receptor for fibrinogen (Fg), von Willebrand factor (wWF), and fibronectin (Fn) (see Phillips et al., Blood (1988) 71:831-843); however, it is assumed that the binding of fibrinogen and/or von Willebrand factor is mainly responsible for blood aggregation and thrombus formation in vivo. Compounds that specifically inhibit the binding of fibrinogen or von Willebrand factor to GP Ilb-IIIa inhibit platelet aggregation and may be candidates for inhibiting thrombus formation in vivo.

Platelet GP Ilb-IIIa is now known to be a member of a superfamily of structurally related adhesive protein receptors known collectively as "integrins". Like GP Ilb-Illa, all the integrins known up to now are bisubunit molecules with a larger alpha subunit (eg GP Ilb) and a smaller beta superunit (eg GP Illa). There is a high degree of homology between the known sequences of the integrin subunits, and this suggests that the integrins have evolved from a common precursor. Integrins act in various cellular adhesions, and have been found in leukocytes, endothelial cells, smooth muscle cells and other cells in the musculature. Because of. that integrins are widely dispersed, while GP IIb-IIIa is restricted to the platelets, a preferred anti-aggregation agent will selectively inhibit GP IIb-IIIa in relation to other integrins.

Several classes of peptides have been described that block the binding of adhesive proteins to activated platelets, and that inhibit platelet aggregation (see Hawiger et al., US Patent 4,661,471; and Rouslahti et al., US Patents 4,614,517; 4,578,079; 4,792,525; and UK application GB 2,207,922A). In one class of peptides, the sequences RGD are critical, and the tetra-peptide sequences RGDS, RGDT, RGDC have been specifically used. The amino acid sequence RGDX is found in various adhesive proteins including Fg, Vn, vWF and Fn. This sequence has been demonstrated to play an important role in the interaction of adhesive proteins with adhesive protein receptors due to that peptides containing these sequences block the binding of adhesive proteins. See e.g. Pierschbacher, M.D. et al., J. Biol. Chem (1987) 26217294-17298; Ruggeri et al., Proe. Nati. Acad. Pollock. (USDA) (1986) 83:5708-5712; and Rouslahti et al. Cell (1986) 44:517-518. Tetrapeptides containing this sequence are described in EP application 319,506 published 7 June 1989. Short peptides containing homoarginine instead of arginine in the RGD sequences are described in PCT application W089/07609 published 24 August 1989.

The structural variations allowed in RGD-containing peptides have been investigated by Pierschbacher M.D. et al., J. Biol. Chem. (supra). In these studies, it was found that manipulation of the RGD-containing sequence not only affected the activity related to inhibition of the binding of fibronectin or vitronectin to the substrate, but could also affect the differentiation between the binding of the two ligands. The peptide sequence GRGDSPC taken from the cell junction domain of fibronectin was used as a model peptide. Certain substitutions, such as replacement of L-Arg by D-Arg appear to have no effect on the binding of either ligand, but substitution of D-Ala for Gly, or D-Asp for L-Asp destroys inhibitory activity- ten. Because of. that substituting D-Ser for L-Ser reduced the inhibition of nitronectin interaction with the vitronectin receptor, there was little effect on fibronectin interaction with the fibronectin receptor; substitution of Asn for Ser resulted in a peptide that had enhanced inhibition of fibronectin binding, and a reduced effect on vitronectin binding. Alternative substitutions for Ser had other effects. Threonine substituted for Ser gave a peptide with increased inhibition for binding to the nitronectin receptor and substitution of L-Pro led to an inactive peptide. A cyclic peptide was also prepared of the sequence Gly-Pen-Gly-Arg-Gly-Asp-Ser-Pro-Cys-Ala, where "Pen" is penicillamine, and a disulfide bridge was formed between Pen and Cys. According to the authors, penicillamine had a function that led to an increase in conformational inhibition on the ring, while N-terminal Gly and carboxy-terminal Ala were added to distance the free amino and carboxyl groups from the ring. This cyclic peptide could inhibit nitronectin binding more strongly than the same peptide before cyclization, but was ineffective in inhibiting fibronectin binding.

Recently, an antithrombotic peptide with a modification of the RGD sequence having the "R" residue alkylated was reported by Samanen, J. et al., J Cell Biochem (1990) Suppl 14A:A229. An overview of structure/activity relationships in RGD-containing peptides has been published by Ali, F.E. et al. in Proe. llth Am. Peptide Symp., Marshall et al., ed. ESCOM Leiden 1990. European patent application, publ. No. 341,915 published November 15, 1989 discloses two groups of peptides, one linear and the other cyclic, that bind the platelet GP IIb-IIa receptor and thus to inhibit its ability to bind vWF, fibronectin and fibronogen-f ibrin. No data have been provided relating to the specificity of binding of these peptides. The group of cyclic peptides includes modifications of the RGD sequence in which R is substituted with D or L homarginine, dimethyl or diethylarginine, lysine or an alpha-alkylated derivative of these residues. Minimal cyclic structures comprise only the "R" GD sequence sandwiched between the two residues that form the disulfide bridge.

A separate class of inhibitory peptides utilize peptide sequences shown on the carboxyl-terminal sequence derived from the gamma chain of fibrinogen, dodecapeptide HHLGGAQKAGDV (Kloczewiak et al., Biochemistry (1989) 28:2915-2919; Timmons et al., (Ibid), 2919-2923 US Patent 4,661,471 (supra); EP Application 298,820). Although this sequence inhibits Fg and vWF binding to GP Ilb-IIIa and subsequent platelet aggregation, the utility of this peptide is limited due to that it has a low affinity for interaction with the blood-platelet receptors (IC50= 10-100 uM).

Recently, several groups have isolated and characterized a new class of low molecular weight polypeptide factors from snake venoms that have extremely high affinity for the GP Ilb-IIIa complex. Huang, T.-F., et al., J. Biol Chem (1987) 262:16157-16163; Huang, T.-F., et al., Biochemistry (1989) 28:661-666 reports the primary structure of trigramin, a 72 amino acid peptide containing RGD and 6 disulfide bridges isolated from Trimeresurus gramineus. Gan, Z.-R., et al., J. Biol. Chem. (1988) 263:19827-19832 report the properties and structure of echistatin, a 49 amino acid peptide also containing RGD and 4 putative disulfide bridges isolated from Echis carinatus. Williams, J.A. et al., FASEB Journal

(1989) 3:A310, Abstr. No. 487m, reports the sequence and properties of the related peptides elegantln, albolabrin and flavovlrldln. In addition, characterization of bitistatin was reported by Shebuski, R. J. et al., J. Biol. Chem. (1989) 264:21550-21556; and PAI from Agkistrodon piscivorus piscivorus reported by Chao, B. H., et al., Proe. Nati. Acad. Pollock. USA

(1989) 86:8050-8054. The relationship between different GP IIb-Illa antagonists from snake venoms was discussed by Dennis, M.S., et al., Proe. Nati. Acad. Pollock. USA (1989) 87:2471-2475.

Included in this group of inhibitory peptides from snake venoms are alboadrin isolated from Trimeresurus albo-labris, elegatin isolated from T. elegans, flavoviridin isolated from T. flavoviridis, bactroxostatin isolated from Bothrops atrox, bistatin isolated from Bitis arietans reported by Niewiarowski, S., et al., Throm's Haemostas (1989) 62:319 (Abstr. SY-XIV-5). In addition, applagin has been purified from Agkistrodon p. piscivorus and reported by Chao, B. et al., Throms. Haemostas (1989) 62:50 (Abstr. 120) and halysin, purified from Agkistrodon halys as reported by Huang, T. F. et al., Throms haemostas (1989) 62:48 (Abstr. 112). All these peptides show a high degree of sequence homology. All the peptides reported up to now from snake venoms that inhibit the binding of adhesive proteins to integrin receptors contain the RGD sequence.

Despite the fact that these reported snake venom factors are potent inhibitors of platelet aggregation in vitro, these peptides also bind with high affinity to other members of adhesive protein receptors such as vitronectin and fibronectin receptors (Knudsen, K. A. et al., Exp. Cell. Res. (1988) 179:42-49; Rucinski, B. et al., Throm's Haemostas

(1989) 62:50 (Abstr. 120). This lack of specificity for snake venom factors for GP Ilb-IIIa is an undesirable feature of their therapeutic use as inhibitors of thrombus formation, because that they have the potential to provide the adhesive properties to other cells in the vasculature, especially the adhesions mediated by integrins.

Another approach developed for the formation of platelet thrombus inhibitors has been the use of murine anti-GP IIb-Illa monoclonal antibodies that block the binding of the adhesive proteins to stimulate platelets. These monoclonal antibodies have been used to prevent coronary arteriole occlusion after reperfusion with a tissue plasminogen activator in dogs (Yasuda, T. et al., J. Clin. Invest. (1988) 81:1284-1291) and prevent cyclic reduction of flow in injured rabbit coronary arteries with a high degree of stenosis. Potential side effects of using such monoclonal antibodies in humans may result from their long-acting effects and from their potential immunogenicity.

Additional therapeutic treatment regimens are needed to prevent or at least mitigate unwanted thrombus formation. Therapeutic agents that can block or inhibit thrombus formation at specific locations without compromising hemostasis and without affecting other cellular adhesions will provide great therapeutic benefits. These agents should be potent, specific for GP IIb-IIIa, and nonimmunogenic to most patients, and should also be easy to administer, stable, and economical to manufacture. These agents should act transiently, and could act on the earliest stages of thrombus formation without interfering with long-term hemostasis.

The present invention therefore relates to an analogue method for the production of a therapeutically active specific platelet aggregation inhibitor (PAI) peptide with the ability to inhibit the binding of Fg or vWF, GP Ilb-IIIa with substantially more, that is to say that either the percentage inhibition is at least two times higher at a given concentration of PAI or that the concentration of PAI causing 50$ inhibition is at least twofold less for the Fg or vWF/GP Ilb-IIIa binding inhibition than for alternative ligand/receptor binding, i.e. potency than inhibition of the binding of vitronectin to the vitronectin receptor or fibronectin to the fibronectin receptor, where PAI has the formula:

in which K<*>is a lysyl residue of the formula:

R<1>2N(CH2)4CHNHC0- ,

wherein each R<1> is independently H, alkyl (1-6C) or at least one R<1> is R<2->C=NR3 ,

wherein R<2> is H, alkyl (1-6C) or is a substituted or unsubstituted phenyl or benzyl residue, or NR<4>2where each R<4> is independently H or alkyl (1-6C), and

R<3> is H, alkyl (1-6C), phenyl or benzyl, or r<2->C=NR<3> is a residue selected from the group consisting of:

where m is a number of 2-3, and each R 1 is independently H or alkyl (1-6C);

and wherein one or two (CH 2 ) may be replaced by 0 or S provided that said 0 or S is not adjacent to another heteroatom;

AA is a small, neutral (polar or nonpolar) amino acid, and ni is a number from 0-3;

AA2 is a neutral, non-polar, large (aromatic or non-aromatic) or a polar aromatic amino acid, and n2 is a number of 0-3;

AÅ3 is a proline residue or a modified proline residue and n3 is a number of 0-1;

AA4 is a neutral, small amino acid or the N-acylated form thereof, and n4 is a number of 0-3;

each of and X 2 is independently a residue capable of forming a bond between X 1 and X 2 to achieve a cyclic compound as shown;

each of Y 1 and Y 2 is independently a non-interfering substituent, or may be absent;

where one or more peptide bonds can optionally be replaced by a bond selected from the group consisting of —CH2NE—, -CH2S-, CH2CE2-, -CE=CH- (cis and trans), -C0CH2-, -CH(OH)CH2- and -CH2SO-;

provided that if n3 is 0; is either:

1) the sum of n2 and n4 at least 2; or

2) K<*>must be different from Har or K; or

3) X2 must be different from Cys (C), penicillamine (Pen), or 2-amino-3,3-cyclopentanemethylene-3-mercaptopropionic acid (APmp); or 4) one or more peptide bonds are replaced with said alternating bond, characterized in that it

includes:

starting the synthesis from the carboxy-terminal end of said peptide using an amino acid having an alpha-amln protecting group, in which said amino acid is attached to a resin support, completing the formation of a protected peptide-resin chain, removing said protecting group, cleaving the said peptide from said resin, cyclization of said peptide and optionally purification of said peptide.

The understanding that platelet aggregation is mainly achieved through the binding of fibrinogen and/or vWF to GP Ilb-IIIa on the surface of platelets when the platelets are treated with appropriate stimuli, such as ADP, is important. When applying these criteria, i.e. inhibition of binding of fibrinogen and/or vWF to isolated receptor and analogous criteria related to inhibition of binding of fibronectin (Fn) to the fibronectin receptor (Fn/FnR binding) and vitronectin to the vitronectin receptor (Vn/VnR -binding), as well as binding of other factors, such as Fn and Vn to GP IIb-Illa, a specificity profile for the platelet aggregation inhibitor (PAI) can be quickly and conveniently obtained. This approach has been used to screen and characterize an extensive panel of snake venoms for the presence or absence of PAIs, to characterize the specificity of PAIs identified from this panel on their specificity to inhibit binding to GP Ilb-IIIa relative to inhibition of other integrins, and to identify active peptides that are derivatives of PAI.

PAI can be isolated from active snake venom and from Echis colorata, Eristicophis macmahonii; A. hypnale, A. acutus, A. piscivorous leucostoma, A. piscivorus conanti; Bothrops asper; Bothrops cotiara, B. jararaca, B. jararacussi, B. lansbergi, B. medusa, B. nasuta, B. neuwiedi, B. pradoi, B. schlegli, Crotalus atrox, C. basilicus, C. cerastes cerastes, C. durissus surissus, C. durissus totonatacus, C. horridus horridus, C. molossus molossus, C. ruber ruber, C. viridis cereberus, Crotalus v. helleri, Crotalus v. lutosus, Crotalus v. oreganus, Crotalus v. viridis; Lachesis mutas; Sistrurus catenatus tergeminus, and Sistrurus milarus barbouri.

PAI in isolated form prepared from, or having the amino acid sequences of, those obtained from Eristicophis macmahonii (eristicophin); Bothrops cotiara (cotiarin); B. jararacussu; Crotalus atrox (crotatroxin); Crotalus basilicus (basilicin); C. cerastes cerastes (cerastine); C. durissus totanatacus (durissin); C. durissus durissus (durissin); C. h. horridus (horridin); Crotalus m. molossus (molossin); C. ruber ruber (ruberine); Crotalus viridis lutosus (lutosin); C. v. viridis (viridin); Crotalus v. oreganus (oreganin); Crotalus v. helleri; Lachesis mutas (lachesine); Sistrurus catenatus tergeminus (tergeminin); and S. milarus barbouri (barbourin) are preferred.

Particularly preferred are eristicophin, cotiarin, crotatroxin, cerastin, durissin, horridin, ruberin, lachesin, basilicin, lutosin, molossin, oreganin, viridin, tergeminin and barbourin.

In another main aspect, the invention is directed to the production of a group of peptides or peptide-related compounds which are generally platelet aggregation inhibitors which can inhibit the binding of Fg or vWF to GP Ilb-IIIa at a significantly higher potency than that where it inhibits the binding of vitronectin to the vitronectin receptor or fibronectin to the fibronectin receptor. These peptides are characterized by having the binding sequence K<*>GDX instead of the RGDX binding sequence found in earlier PAI proteins. Barbourin PAI isolated from S. milarus barbouri is an illustration of this series of peptides. However, shorter forms of this peptide can also be used, as well as analogous sequences that also contain 1-10 amino acid residue modifications elsewhere in the peptide chain, and/or replacement of peptide bonds with alternative bonds. Other illustrative embodiments include isolated PAI peptides with a native RGDX sequence where this has been replaced with K<*>GDX. As in the case of barbourin, these isolated PAIs may be in native form or may be shortened and/or contain 1-10 amino acid residue substitutions or deletions, and/or may have non-peptide bonds substituted for peptide bonds.

Another illustrative group of embodiments are peptides or modified peptides with a specific PAI activity of formula:

where K<*>is a substituted or unsubstituted lysyl residue of formula R<1>2N(CH2)4CHNHCO- as described above,

where each R<1> is independently H, alkyl (1-6C) or at most one R<1> is R<2->C=NR<3>,

where R<2> is H, alkyl (1-6C) or a substituted or unsubstituted phenyl or benzyl residue, or NR<4>2where each R<4> is independently H or alkyl(1-6C), and

R<3> is H, alkyl(1-6C), phenyl or benzyl, or

R<2->C=NR<3>is a residue selected from the group consisting of:

where m is a number of 2-3, and each R<5> is independently H or alkyl(1-6C);

and wherein one or two (CH 2 ) may be replaced by 0 or S provided that said 0 or S is not adjacent to another heteroatom;

AA-1 is a small, neutral (polar or nonpolar) amino acid, and ni is a number from 0-3;

AA2 is a neutral, non-polar, large (aromatic or non-aromatic) or a polar aromatic amino acid, and n2 is a number of 0-3;

AA3 is a proline residue or a modified proline residue (as defined below), and n3 is a number of 0-1;

AA4 is a neutral, small amino acid or n-alkylated form thereof, and n4 is a number of 0-3;

each of and X 2 is independently a residue capable of forming a bond between X 1 and X 2 to achieve a cyclic compound as shown; and

each of and Y 2 is independently a non-interfering substituent, or may be absent;

wherein one or more peptide bonds may optionally be replaced by a bond selected from the group consisting of —CH2NH-, -CH2S-, -CH=CH- (cis and trans), -COCH2-, -CH(OH)CH2- and -CH2S0 -;

provided that if n3 is 0; either:

1) the sum of n2 and n4 must be at least 2; or

2) KH must be different from Har or K; or

3) X2 must be different from cys (C), penicillamine (Pen) or 2-amino-3,3-cyclopentanemethylene-3-mercaptopropionic acid (APmo); or 4) Y±or Y2 must comprise at least one amino acid residue; or 5) one or more peptide bonds are replaced with said alternative bond.

Other aspects of the invention relate to recombinant methods and materials for the production of these and other related peptides, methods for in vitro production thereof, pharmaceutical compositions containing these compounds, and methods for inhibiting platelet aggregation and thrombus formation using these compounds and compositions.

Brief description of the drawings:

Fig. 1 shows inhibition of the binding of fibrinogen to GP IIb-IIIa by partially purified snake venoms. Fig. 2 shows the dose-response adhesion inhibition of Centricon-10 ultrafiltrates of crude venoms in both fibrinogen/GP IIb-IIIa and vitronectin/vitronectin receptor assays. Fig. 3 shows the HPLC profile of crude PAI from Eristicophis macmahoni venom. The shaded area contains the biologically active fractions. Fig. 4 shows the HPLC profile of the PAI fractions from fig. 3. The shaded area contains the bioactive fractions. Fig. 5 shows the analytical HPLC profile of the PAI fractions from fig. 4. Fig. 6 shows the complete amino acid sequences of eristicophin, barbourin, tergeminin, cerastin, ruberin, lachesin, cotiarin, crotatroxin, horridin. lutosin, viridin, molossin, basilicin, durissin, jararacin, cerebrin and oreganin, and enzyme-cleaved fragments determined by automated Edman degradation. Fig. 7 shows the HPLC profile of PAI obtained from the G-50 fractions of crude Sistrurus c. tergeminus venom. Fig. 8 shows the HPLC profile of the PAI fractions from fig. 7. Fig. 9 shows the activity of purified PAI according to fig. 8 in inhibitory binding in several receptor assays. Fig. 10 shows the HPLC profile of the platelet aggregation inhibitor obtained from the G-50 fractions of crude Sistrurus m. barbouri friend. The shaded areas contain the bioactive fractions. Fig. 11 shows the HPLC profile of active PAI fractions according to fig. 10. Fig. 12 compares the amino acid sequences of a number of PAIs relative to barbourin. Fig. 13 shows the HPLC profile of crude PAI from Lachesis mutas venom. Shaded areas contain the biologically active fractions. Fig. 14 shows the HPLC profile of PAI-active fractions from fig. 13. The shaded area contains the biologically active fractions. Fig. 15 shows the analytical HPLC profile of the PAI fractions according to fig. 14 from Lachesis mutas. Fig. 16 shows the HPLC profile of crude PAI from Crotalus viridis viridis venom. The shaded area contains the biologically active fractions. Fig. 17 shows the HPLC profile of the PAI fractions according to fig. 16. Fig. 18 shows the dose-response effects of purified snake venom peptides for inhibited fibrinogen/GP Ilb-IIIa binding compared to echistatin. Fig. 19 shows the dose-response effects of purified snake venom peptides for inhibiting ADP (4 µM) induced human platelet aggregation in platelet-rich plasma (PRP), compared to echistatin. Fig. 20 shows the activity profile from HPLC fractionation of C.

c. cerastes venom.

Fig. 21 shows the results of the HPLC analysis of active fractions from fig. 20. Fig. 22 shows the activity profile of HPLC fractionation of PAI from C. ruber ruber. Fig. 23 shows the activity profile from an analytical C-18 column on homogeneous peptide obtained from C. atrox. Fig. 24 shows the analytical HPLC profile of the homogeneous peptide isolated from Bothrops cotiara. Fig. 25 shows the dose-response effects of purified cotiarin on inhibition of the binding of fibrinogen to GP IIb-IIIa and inhibition of the binding of vitronectin to the vitronectin receptor. Fig. 26 shows the effects of purified snake venom peptides on the binding of fibrinogen to GP IIb-IIIa and vitronectin to the vitronectin receptor. Fig. 27 shows the results of the binding activity of analog No. 1, [E<2>8L41C64 )barbourin (28-73) with respect to GP IIb-IIa and the vitronectin receptor. Fig. 28 shows the ability of the synthetic eristicophin analogue to inhibit the binding of fibrinogen to GP IIb-IIIa and the inability to inhibit the binding of vitronectin to the vitronectin receptor. Fig. 29 shows the ability of linear and cyclic RGDW compounds and linear and cyclic KGDW compounds to inhibit the binding of fibrinogen to GP IIb-IIIa. Fig. 30 shows the ability of different KGDW analogs to inhibit the binding of fibrinogen to GP IIb-IIIa and to inhibit the binding of vitronectin to the vitronectin receptor. Fig. 31 shows the ability of various native and synthetic platelet aggregation inhibitors to inhibit the attachment of M21 melanoma cells to vitronectin. Fig. 32 shows the ability of RGDS and a cyclic RGD compound to inhibit the attachment of M21 melanoma cells to vitronectin and the inability of a cyclic KGDW analog to inhibit the attachment of M21 melanoma cells to vitronectin. Fig. 33 shows the activity of analog No. 60, Mpr-(Har)-G-D-W-P-C-NHg, for inhibiting platelet aggregation and cell adhesion to vitronectin. Fig. 34 shows the activities according to fig. 33 for analog 19, Mpr-K-G-D-W-P-C-NH2. Fig. 35 shows initiation of cyclic flow reductions (CFRs) in an open chest dog model of thrombosis (Folt's model). Fig. 36 shows the effects of a 10 mg bolus dose administration on CFRs initiated in the open chest dog model of thrombosis (Folt's model). Fig. 37 shows the effects of a 40 mg bolus dose administration on CFRs initiated in the open chest dog model of thrombosis (Folt's model). Fig. 38 shows the full-length DNA sequence encoding the amino acid sequence of barbourin (1-73). Fig. 39 shows the DNA sequence coding for [M-1L41]barbourin (1-73) ligated to the PhoA leader sequence. Fig. 40 shows the DNA sequence encoding analog No. 1 linked to a PhoA sequence for expression in bacteria. Fig. 41 shows oligonucleotides used in a PCR reaction to obtain DNA-encoding analog No. 1. The amino acids included in the analog per se are shown in bold. Fig. 42 shows the border sequence of the tandem repeats of analog No. 1 coding DNA. Fig. 43 shows a diagram of the shortened barbour ring as tandem repeats.

Methods of carrying out the invention

The invention provides platelet aggregation inhibitors (PAIs) that can be isolated from snake venom that have been identified as active by an assay method and compounds that have similar structures and are prepared using standard in vitro techniques, such as solid phase peptide synthesis or using recombinant methods, or combination of these. Some of these inhibitors are uniquely specific for inhibition of platelet aggregation, and do not inhibit alternative binding within the integrin family. Others have different specificity areas. The isolation of naturally occurring PAI from snake venom is described below; construction of inhibitors which have significantly higher potency in terms of inhibition of platelet aggregation than inhibition of e.g. the vitronectin/- vitronectin receptor interaction by incorporating a K*GD sequence relative to RGD; method for the production of these peptides; method of recombinant production; antibodies raised against the peptides prepared according to the invention; the assay procedure enabling the identification of snake venoms containing PAI; and administration and application of PAI.

PAI means a factor that can prevent the aggregation of stimulated platelets in standard analyses, e.g. those described by Gan, Z.R., et al., and Huang, T.-F., et al., (supra). In these analyses, washed platelets are combined with fibrinogen, Ca<+2> and the material being tested. The platelets are stimulated with ADP (or other known stimulants or combinations thereof), and aggregation (or lack thereof) is observed by e.g. application of a commercially available aggregometer.

Some of the PAIs prepared according to the invention have been identified as specific for the inhibition of the binding of fibrinogen and/or vWF to GP IIb-IIIa. It appears that the specificity is of different degrees, and therefore a PAI specific for inhibition of Fg or vWF binding to GP Ilb-IIIa inhibits this binding significantly more than it inhibits the binding of Fn to FnR, or Vn to VnR. By "significantly more" is meant that either the percent inhibition is at least twofold higher at a given concentration of PAI or that the concentration of PAI causing 50% inhibition is at least twofold less for Fg or vWF/GP Ilb-IIIa binding inhibition than for alternative ligand/- receptor binding.

Isolated native PAI and purification methods

Platelet aggregation inhibitors (PAIs) include low molecular weight peptides which can be prepared in isolated form, as described below, from snake venom which has been identified as "active", i.e. found to contain PAIs.

It is possible with a simple identification and characterization of the presence of an effective PAI in snake venom that selectively inhibits the binding to GP Ilb-IIIa in relation to other integrins such as, e.g. the vitronectin receptor and the f ibronectin receptor. Upon such identification and possibly and optimally, characterization, PAI can be isolated and purified using various standard techniques illustrated herein and described within the field. A combination of separation based on molecular weight (usually isolation of compounds of < 10 kd), ion exchange chromatography and reverse phase HPLC can be used. Other techniques may also be used, but a suitable procedure applicable to PAI from any active snake venom is as follows: Approximately 10-1000 mg of venom is dissolved in dilute acetic acid and applied to a sizing column, such as Sephadex G-50, and eluted in the same solvent. The fractions are assayed for activity using the Fg/GP Ilb-IIIa binding assay of the invention, a standard blood platelet aggregation assay (PAA) or any similar assay based on the adhesive protein binding activity of GP Ilb-IIIa. Alternatively, the fraction of < 10 kd according to the fraction of the venom can be isolated using ultrafiltration and analyzed in a similar manner.

Low MW fraction isolated by either method was then applied to a preparative C-18 HPLC column, such as a C-18 Delta Pak reverse phase HPLC column, available from Waters, pre-equilibrated in 0.1% trifluoroacetic acid (TFA )/8# acetonitrile. Adsorbed PAI was then eluted using a gradient of 8%-b% acetonitrile in 0.1% TFA. The slope of the gradient and the flow rate are optimized using routine procedures. Active fractions are determined by PAI or by receptor-binding method. The active fractions were then pooled, concentrated and tested for homogeneity using analytical HPLC or SDS-PAGE. Additional reverse-phase HPLC gradient purification is applied until isolated PAI is homogeneous.

PAIs produced according to the invention, and consequently obtainable by the foregoing or other purification methods, include those from venoms selected from the group consisting of Echis colorata, Eristicophis macmahonii; A. hypnale, A. acutus, A. piscivorous leucostoma, A. piscivorus conanti; Bothrops asper; Bothrops cotiara, B. jararaca, B. jararacussu, B. lansbergi, B. medusa, B. nasuta, B. neuwiedi, B. pradoi, B. schlegli; Crotalus atrox, C. basilicus, C. cerastes cerastes, C. durissus durissus, C. durissus tononatacus, C. horridus horridus, C. molossus molossus, C. ruber ruber, C. viridis cereberus, Crotalus v. helleri, Crotalus v. lutosus, Crotalus v. oreganus, Crotalus v. viridis; Lachesis mutas; Sistrurus catenatus tergeminus, and Sistrurus milarus barbouri.

Preferably, the PAI in isolated form is prepared from, or with the amino acid sequences of, those obtained from Eristicophis macmahonii (eristicofin); Bothrops cotiara (cotiarin); B. jararacussu; Crotalus atrox (crotatroxin); Crotalus basilicus (basilicin); C. cerastes cerastes (cerastine); C. durissus totonatacus (durissin); Crotalus d. durissus (durissin); C. h. horridus (horridin); Crotalus m. molossus (molossin); C. ruber ruber (ruberine); Crotalus viridis lutosus (lutosin); Resume viridis (viridin); Crotalus v. oreganus (oreganin); Crotalus v. helleri; Lachesis mutas (lachesine); Sistrurus catenatus tergeminus (tergeminin); and S. milarus barbouri (barbourine). Particularly preferred is the PAI specific for inhibition of Fg or vWF/GP Ilb-IIIa binding, e.g. those from Sistrurus m. barbouri.

Particularly preferred are eristicophin, cotiarin, crotatroxin, cerastin, durissin, horridin, ruberin, lachesin, basilicin, lutosin, molossin, oreganin, viridin, tergeminin and barbourin.

Purified PAI can be sequenced using standard procedures thus enabling synthesis using standard solid phase techniques (especially for shorter forms of PAI) or recombinant production. E.g. an Applied Biosystems Sequenator can be used according to carboxyamido-methylation or pyridylethylation of the peptide as described by Huang et al., J. Biol. Chem (1987) 262:16157-16163 followed by desalting the sample on a C-18 Delta Pak column using 0.1% TFA and acetonitrile.

Isolated PAI with specific sequence can, when produced in vitro, be modified by frequency alterations that do not destroy the activity. In general, these modified forms will differ from the native forms by 1-10, preferably 1-4, amino acid substitutions or will be in shortened forms. In addition, one or more peptide bonds may be replaced with alternative bonds as described below. In particular, substitution by replacing RGD with K<*>GD to provide GP Ilb-IIIa specificity as described below.

The PAI of Sistrurus m. barbouri has been purified to homogeneity and sequenced, and designated "barbourin". Unlike the adhesive proteins for GP Ilb-IIIa identified so far and the peptides from snake venoms that block GP Ilb-IIIa function, barbourin does not contain the standard Arg-Lys-Asp sequence of the adhesive proteins known in this area. The apparent binding sequence in barbourin is Lys-Gly-Asp-(Trp). Presence of the KGD sequence in the apparent binding region of this peptide is particularly surprising in light of the observation that substitution of Lys for Arg in small synthetic peptides based on the RDGX sequence reduces the ability of these peptides to bind to the integrin receptors (Pierschbacher et al., Proe. Nat. Acad. Sei (USA) (1984) 81:5985-5988; Williams et al., Thromb Res (1987) 46:457-471); Huang et al., J. Biol. Chem.

(1987) 262:16157-16163. It is believed that this substitution may be partially responsible for the specificity of the barbourin peptide for inhibiting Fg and vWF binding to GP Ilb-IIIa, in contrast to e.g. inhibition of vitronectin binding to the vitronectin receptor.

K«GDX-containing peptides

The "Barbourin" peptide has been shown to have the binding sequence KGDX relative to RGDX found in the PAI compounds of the prior art. Presence of KGDX in this PAI sequence appears to be associated with a preferential affinity for GP IIb-IIIa over vitronectin or the fibronectin receptors. The effect of the substitution of a lysyl residue for an arginine in the frequency appears to be related to the increased length of the side chain together with maintained basicity of the nitrogen as further described below. It appears that it is not the lysyl residue per se that accounts for the increased activity and specificity, but rather the region provided by this homologous extension of substituted arginine. The peptides containing K<*>GDX in the binding sequence are significantly more potent in inhibiting the binding of Fg or vWF to GP Ilb-IIIa compared to their ability to inhibit the binding of vitronectin to the vitronectin receptor and the binding of fibronectin to the fibronectin receptor. As stated above, "significantly more" potent for inhibiting the preferred binding means that percent inhibition is at least twice as great at a given concentration of inhibitor or that the concentration of PAI causing 50% inhibition is at least 2-3 times less for binding of Fg or vWF to GP Ilb-IIIa than for binding of other ligands to other integrins.

"Alkyl" is conveniently defined as a linear or branched chain or cyclic heterocarbyl radical with a specified number of carbon atoms such as methyl, ethyl, isopropyl, N-hexyl, 2-methylbutyl, cyclohexyl, etc.

Benzyl and phenyl residues represented by R<2> can be unsubstituted, or be substituted with non-interfering substituents. Preferred substitution patterns are those in which only one substituent is attached to the aromatic nucleus, preferably in the 4-position. Preferred substituents are electron-donating substituents such as alkyl, especially ethyl or methyl, or phenyl.

Examples of fragments and/or modified forms of naturally occurring snake venom PAI include [E<28>,L<41>,C^<4>]bar-bourin (28-73) with the sequence: and [K<2>^] eristicophin (4-51) of the sequence.

In this description, the size of the fragment is indicated in parentheses after the name with the numbers of the amino acids that were included in the fragment, and the prefix letters in parentheses and the numbers indicate amino acid substitutions at the numbered positions in the native full-length peptide. For the barbourin fragment above, the length of the fragment covers residues 28-73 inclusive of the native sequence and the amino acids originally at positions 28, 41 and 64 of the numbered native sequence which have been replaced with Glu (E), Leu (L) and Cys (C).

As further examples, the arginine of the RGD sequence appearing in trigramin, elegantin, albolabrin, crotatroxin, flavoviridin, echistatin, bitistatin, viridin, molossin, lutosin, basilicin, applagin, halysin, horridin, tergeminin, lachesin, cotiarin, cereberin, jararacin, kistrin, eristicofin, bitan-a and ruberin/oreganin can be substituted with a K' residue to provide specifically active PAIs with a preferred affinity for GP IIb-IIIa. In addition, shortened forms of these peptides, containing at least 20, preferably at least 30, and more preferably at least 40 amino acids, can be prepared from the native peptide or in this modified form. In addition, or in the alternative, 1-10, preferably 1-4 amino acids irrelevant to the RGD/K<*>GD sequence may be substituted or modified, preferably with conservative amino acid substitutions. By conservative amino acid substitutions is meant e.g. substitution of an acidic amino acid residue for an acidic amino acid residue, neutral for neutral, basic for basic, etc. as further described below.

Y± and Y2 can be peptide extensions of 0-25 amino acid residues and can be in derivatized form. The Y^N terminal extension can e.g. be acetylated or otherwise acylated; The Y2C terminal extension can be amidated with NH2 or with a primary or secondary amine of the formula R-NH2 or R2NH wherein each R is independently a lower alkyl of 1-4C such as methyl, n-butyl or t-butyl. Y^ can also be (H) or acyl; Y can be (OH), NH 2 or an amine as above. Where the compound of formula (1) is a single cyclic peptide, Y1 and Y2 are absent.

Xi and X2 are usually amino acid residues capable of cyclization such as e.g. and most preferably cysteine residues which can form a disulfide ring. Other residues which can form disulfide or other bonds can also be used - e.g. Pen (penicillamine)—the residue described by Pierschbacher et al. (supra) or Mpr (mercarptopropionyl) or Mvl (mercaptovaleryl)—the residue. Other types of covalent bonds for cyclization that may be considered include peptide bonds such as an amide formed between the side-chain amino group of a lysyl residue with a side-chain carboxyl group of a glutamyl residue and ester bonds, which will be formed between a side-chain alcohol of a threonine residue with a side-chain carboxyl of an aspartyl residue. Any compatible residue capable of forming peptide bonds with the remainder of the chain (or modified peptide bonds as described above) and capable of forming covalent bonds to achieve cyclization may be used. This includes e.g. simple cyclic peptides, in which a peptide bond is directly formed between NH2 and the N-terminus and COOH at the C-terminus.

As described above, one or more of the indicated peptide bonds may be replaced by a substitute bond such as -CH2NH-, -CH2-S-, CH2CH2-, -CH=CH- (cis and trans), -COCH2-,

-CH(OH)CH2- and -CH2SO-.

In the designation of the amino acid residues AA1-AA4 above, the description has been formed on the basis of a specification method in which the amino acid residues can generally be subclassified into four main subclasses. This classification is also shown in diagram form below.

Acidic: The residue has a negative charge due to loss of H-ion at physiological pH, and the residue is attracted by aqueous solution to seek surface positions in the conformation of a peptide where it is included when the peptide is in aqueous medium at a physiological pH.

Basic: The residue has a positive charge due to the association with the H ion at a physiological pH, and the residue is attracted to aqueous solution to seek surface positions in the conformation of a peptide where it is included when the peptide is in aqueous medium at a physiological pH.

Neutral/nonpolar: The residues are uncharged at physiological pH, and the residue is repelled by aqueous solution to seek the internal positions in the conformation of the peptide in which it is contained when the peptide is in aqueous medium. These residues are also termed "hydrophobic" herein.

Neutral/polar: The residues are not charged at a physiological pH, but the residue is attracted to aqueous solution to seek the outer positions in the conformation of a peptide in which it is contained when the peptide is in aqueous medium.

It goes without saying that there is a statistical collection of individual residue molecules where some molecules are charged and some are not, and there will be an attraction for or repulsion from an aqueous medium to a greater or lesser extent. To meet the definition of "charged", a significant percentage (at least about 25%) of the individual molecules are charged at a physiological pH. The degree of attraction or repulsion required for classification as polar or nonpolar is arbitrary, and amino acids specifically contemplated in the invention have been specifically classified as one or the other. Most amino acids that are not specifically named can be classified on the basis of known behavior.

Amino acid residues can further be sub-classified as cyclic or non-cyclic and aromatic or non-aromatic, self-explanatory classifications with respect to the side-chain substituent groups of the residues, and as small or large. The residue is considered small if it contains a total of 4 carbon atoms or less, including the carboxyl carbon. Small residuals are, of course, always non-arornatic.

For the naturally occurring protein amino acids, the sub-classification according to the preceding scheme follows (see also the diagram below).

Acidic: Aspartic acid and glutamic acid;

basic/non-cyclic: arginine, lysine;

basic/cyclic: histidine;

neutral/polar/little: glycine, serine and cysteine;

neutral/polar/large/non-aromatic: threonine, asparagine, glutamine;

neutral/polar/large/aromatic: tyrosine;

neutral/nonpolar/little: alanine;

neutral/non-polar/large/non-aromatic: valine, isoleucine, leucine, methionine;

neutral/non-polar/large/aromatic; phenylalanine and tryptophan.

The gene-encoded amino acid proline, which technically belongs to the groups neutral/apolar/large/cyclic and non-aromatic, is a special case because its known effects on the secondary conformation of the peptide chains, and is therefore not included in this defined group, but is classified separately. AA3 is designated as a proline residue or a "modified proline residue". As is known, proline is a five-membered nitrogen heterocycle with a carboxyl group in the 2-position. Modified proline residues are all nitrogen five- or six-membered heterocycles with carboxyl groups in position alpha to the nitrogen; additional heterocyclic atoms may also be included in the ring. Modified proline residues include residues of pipecolic acid (2-carboxypiperidine, abbreviated Pip) and thiazolidine (Thz). Proline or modified proline residues therefore have the formula:

wherein one or two of the methylene groups may be replaced by NR, S or O, and wherein any ring nitrogen may optionally be substituted with a non-interfering substituent such as alkyl.

These common standing amino acids which are not encoded by the genetic code include e.g. beta-alanine (beta-Ala), or other omega-amino acids, such as 3-aminopropionic, 4-amino-butyric acids, etc., alpha-aminoisobutyric acid (Aib), sarcosine (Sar), Ornithine (Orn), citrulline (Cit ), homoarginine (Har), t-butyl-alanine (t-BuA), t-butylglycine (t-BuG), N-methylisoleucine (N-Melle), phenylglycine (Phg), and cyclohexylalanine (Cha), nor-leucine (Nie), cysteic acid (Cya); pipecolic acid (Pip), thiazolidine (Thz), 2-naphthylalanine (2-Nal) and methionine sulfoxide (MSO). These also appropriately fall into specific categories.

Based on the above definition, Sar and beta-ala are neutral/- apolar/little;

t-BuA, t-BuG, N-Melle, Nie and Cha are neutral/apolar/large/- non-ar ornate in sk;

Har and Orn are basic/noncyclic;

Cya is sour;

Cit, acetyl-Lys and MSO are neutral/polar/large/non-aromatic;

2-Nal and Phg are neutral/nonpolar/large/aromatic; and Pip and Thz are modified proline residues;

The above can be shown diagrammatically as follows:

Amlnosvreklassif ikas. 1 Wed. 1 ema

The various omega amino acids are classified according to size as neutral/nonpolar/small (beta-ala, ie, 3-aminopropynic, 4-aminobutyric) or large (all others).

Other amino acid substitutions for those encoded in the gene may also be included in the peptide compounds within the scope of the invention and may be classified within this general subject matter.

In the formulas representing selected specific embodiments of the present invention, amino and carboxy terminal groups, despite the fact that these are often not specifically shown, will be in the form in which they would be at physiological pH values, if not otherwise is indicated. Thus, N-terminal H<+>g and C-terminal 0~ at physiological pH are present despite the fact that this is not necessarily specified or shown, either in specific examples or in generic formulas. The basic and acidic addition salts, including those formed at non-physiological pH values, are also included in the compounds according to the invention. Unless otherwise stated, the residues are in L-form, and in the generic formulas the specified residues can be either L- or D-. The peptides according to the invention generally have 0, 1 or 2

D residues, preferably 0 or 1, most preferably 0. In the peptides shown, each coded residue is represented where appropriate by a single letter designation corresponding to the trivial name of the amino acid according to the following conventional list:

The amino acids that are not genetically encoded are abbreviated as indicated above.

In the specific peptides shown herein, the L form of any amino acid residue having an optical isomer is intended unless otherwise indicated by a note ("f). Because the residues in the peptides prepared according to the invention are normally in the natural L -optical isomer form, one or two, preferably one, amino acid can be replaced with optical isomer D-form.

Free functional groups, including those at the carboxy or amino terminus, may also be modified by amidation, acylation or other substitution, such as can change the solubility of the compounds without affecting their activity.

When forming amidated peptides, the analogous compounds can be prepared directly, e.g. using the Boc-AAx-pMBHA resin or the Boc-AAx-BHA resin, wherein AAX is the selected carboxy-terminal amino acid of the desired peptide as described below. Alternatively, the peptides produced according to the present invention can be chemically or enzymatically amidated after the peptide production using methods that are well known in the field, or produced by standard solution-phase peptide synthesis protocols.

Certain embodiments of the novo peptides are preferred. In the K<*>(G/Sar)D sequence, G/Sar is preferably G. AA 4 and AA 4 are preferably Gly, Ala or Ser; ni is preferably 0-2, n4 is preferably 1-2. Preferred for AA 2 are neutral/apolar/aromatic amino acids, especially tryptophan and phenylalanine, especially tryptophan, n 2 is preferably 1. X 2 and X 2 are preferably Cys, Mps or Pen (penicillamine) residues. Y-^ is preferably H, acetyl or Gly; Y 2 is preferably -NH 2 or —A-NH 2 . Also generally preferred are C-terminal amidated forms of Y2.

The PAI analogs may include peptides with the following formulas. Despite the fact that all of these have the ability to exhibit cyclic form through the formation of disulfide bonds, these bonds are not shown specifically, and other cyclic forms are denoted by "cyclo".

Preferred peptides

Particularly preferred are the peptides with the formulas:

Klemlian preparation of the peptides

The compounds can be produced chemically using well-known methods such as e.g. solid-phase peptide synthesis. Preparation proceeds from the carboxy-terminal end of the peptide using an alpha-amino protected amino acid, t-butyloxycarbonyl (Boe ) protecting groups can be used for all amino groups although other protecting groups such as fluorenylmethyloxycarbonyl (Fmoc) are suitable. E.g. can Boc-Gly-OE, Boc-Ala-OH, Boc-His (Tos)-OH, (ie selected carboxy-terminal amino acids) be esterified to chloromethylated polystyrene resin supports, p-methylbenzhydrylamine (pMBHA) or PAM resins. The polystyrene resin carrier is preferably a copolymer of styrene with about 0.5 to 2% divinylbenzene as a cross-linking agent which causes the polystyrene polymer to become completely insoluble in certain organic solvents, see Stewart et al., Solid-Phase Peptide Synthesis ( 1969), "W. H. Freeman Co., San Francisco and Merrifield, J. Am. Soc. (1963) 85:2149-2154. These and other methods for preparing the peptide are also exemplified by US patents 3,862,925; 3,842,067 ; 3,972,859; and 4,105,602.

The preparation may use manual synthesis techniques or be applied automatically, e.g. an Applied BioSystems 430A or 431A Peptide Synthesizer (Foster City, California) according to the instructions provided in the instruction manual supplied by the manufacturer. Cleavage of the peptides from the resin can be accomplished using the "low-high" HF cleavage protocols as described in Lu, G.-S. et al., Int. J. Peptide & Protein Res (1987) 29:545-557. Refolding of the snake venom PAI analogs can be accomplished using the methods described in Garsky, V. et al., Proe. Nati. Acad. Pollock. USA (1989) 86:4022-4026 which describes the solid-phase preparation of echistatin.

The cyclic peptides which do not have disulfide bonds can conveniently be prepared by a combination of solid-phase synthesis and formation of cyclic ring structure in solution using the general methods described in US patent 4,612,366 to Nutt. Linear peptides prepared on standard Merridield resin can thus be cleaved from the resin with hydrazine, followed by cyclization of the corresponding acid to form cyclic peptides.

It is obvious to those skilled in the art that the intermediates which are constructed according to the present description during the preparation of the present analog compounds are in themselves new and useful compounds and thus fall within the scope of the invention.

Recombinant production

Alternatively, selected compounds may be prepared by expression of recombinant DNA constructs prepared according to well-known methods. Such production may be desirable to provide large quantities or alternative embodiments of such compounds. Because of. that the peptide sequences are relatively short, the recombinant production becomes easier, but the production by the recombinant methods is particularly preferred in relation to standard solid-phase peptide synthesis for peptides of at least eight amino acid residues.

DNA encoding sequenced PAI is preferably prepared using commercially available nucleic acid synthesis methods. Methods for constructing the expression system for producing PAI in recombinant hosts are also generally known in the art.

Expression can be achieved in either prokaryotic or eukaryotic hosts. Prokaryotes are most often represented by different E. coli strains. Other microbial strains can also be used, such as bacilli, e.g. Bacillus subtilis, various Pseudomonas species or other bacterial strains. In such prokaryotic systems, plasmid vectors containing replication sites and control sequences derived from a species compatible with the host are used. E.g. a working vector for E. coli is pBR322 and its derivatives. Commonly used prokaryotic control sequences containing promoters for transcription initiation, optionally with an operator, together with ribosome binding site sequences, include such commonly used promoters as beta-lactamase (penicillinase) and lactose (lac ) promoter systems, tryptophan (trp) promoter systems and lambda -derived PL promoter and N gene ribosome binding site. Instead, any promoter system compatible with prokaryotes can be used.

The expression systems that have been useful in eukaryotic hosts include promoters derived from appropriate eukaryotic genes. One class of promoters useful in yeast, e.g., includes promoters for the production of glycolytic enzymes, e.g. those for 3-phosphoglycerate kinase. Other yeast promoters include those for the anolase gene or the Leu2 gene obtained from YEpl3.

Suitable mammalian promoters include early and late promoters from SV40 or other viral promoters such as those derived from polyoma adenovirus II, bovine papillomavirus or bovine coma viruses. Suitable viral and mammalian enhancers are listed above. If plant cells are used as an expression system, the nopaline synthesis promoter is, for example, appropriate.

The expression systems are constructed using well-known restriction and ligation techniques and transformed into appropriate hosts.

The transformation is performed using standard techniques appropriate for such cells. The cells containing the expression systems are grown under conditions suitable for the production of PAI, and the PAI is then transformed and purified.

Antibodies

The availability of purified PAI produced according to the invention also enables the production of antibodies specifically immunoreactive with these forms of the active peptide. Compositions containing purified PAI isolated from snake venom or otherwise prepared can be used to stimulate the production of antibodies that immunoreact with the PAI peptide. Standard immunization protocols involving the administration of PAI to various vertebrates, such as rabbits, rats, mice, sheep and chickens, result in anti-sera immunoreactive with the purified peptide. PAI may be conveniently conjugated to a suitable antigenic neutral carrier, such as a suitable serum albumin or keyhole limpet hemocyanin, to enhance immunogenicity. Additionally, the free peptide can be injected with methylated BSA as an alternative to conjugation. Furthermore, antibody-secreting cells of the immunized mammal can be immortalized to prepare monoclonal antibody panels which can then be screened for reactivity with PAI.

The resulting polyclonal or monoclonal antibody preparations are useful in assays for levels of the corresponding PAI in biological samples using standard immunoassay procedures.

Analysis

Identification of the snake venom output material that contains active PAI, and where PAI has known specificity, is possible by analysis. The analysis is based on the observation that compounds that block the binding of fibrinogen to GP Ilb-Illa in vitro can also inhibit thrombin or ADP-induced aggregation of human platelets and the formation of platelet thrombi in vivo. This observation forms the basis for obtaining potent PAI by assessing the ability of the test materials to destroy fibrinogen-GP IIb-IIIa interactions.

In the analysis, GP Ilb-IIIa is produced in purified form, e.g. as described by Fitzgerald. LET. et al., Anal. Biochem.

(1985) 151:169-177, incorporated herein by reference, which are coated onto a solid support such as beads, test tubes or microtiter plates. The coated support is then contacted with fibrinogen and with test material and incubated for a sufficient time to allow maximum binding of fibrinogen to immobilized GP IIb-IIIa. Fibinogen was usually provided at a concentration of approximately 5-50 nm, and the test material can, if desired, be added to a series of dilutions. Usual incubations are 2-4 hours at 35°C, where the time and temperature are dependent on each other.

After the incubation, the solution containing fibrinogen and the test material is removed, and the level of binding of fibrinogen is measured by quantification of bound fibrinogen to GP IIb-IIIa. Any suitable means of detection may be used, but it is convenient to use labeled fibrinogen, e.g. by using radioactive, fluorescing or biotinylated markers. Such methods are well known and need not be described herein.

Assessment of the results is made easier by using a control sample, usually identical to the test sample with the exception that the test sample is absent. In this case, percent inhibition can be calculated using the amount of Fg bound in the control as a basis so that

Other measures of inhibition efficiency, such as ICgg, can also be used.

The analysis systems further comprise characterization of the PAI specificity by binding inhibition assays identical to those above, but substitution of other adhesive proteins for Fg and other receptors for Fg Ilb-IIIa. Inhibition of the binding of vitronectin to the vitronectin receptor; fibronectin to the fibronectin receptor; fibronectin to GP Ilb-IIIa and fibrinogen and/or vWF to GP Ilb-IIIa may be particularly considered. The adhesive protein and receptors for these assays are available within the art.

Other analyses

In addition to the platelet assays, other assays for platelet aggregation inhibitory activity and related activities are also available, as indicated above. A list of commonly used assays is as follows: 1. Plate assays using specific receptors described in previous paragraphs; 2. standard assays directly applied to platelet aggregation, such as those described by Gann, Z.R., et al., J. Biol. Chem. (1988) 263:19827-19832; Huang, T.F. et al., J. Biol Chem (1987) 262:16157-16163; Biochemistry (1989) 28:661-666, cited above and incorporated herein; 3. an in vivo thrombosis model in dogs as described below in Example 1 and by Folts, J.D. et al., Circulation (1976) 54:365; and 4. effect on cell adhesion using S35 methionine-labeled cells as described below in Example 19.

Administration and application

The PAI produced according to the invention is useful therapeutically to prevent thrombus formation. Indications suitable for such treatment include, without limitation, atherosclerosis and arteriosclerosis, acute myocardial infarction, chronic unstable angina, transient ischemic attacks and strokes, peripheral vascular disease, arterial thrombosis, preeclampsia, embolism, restenosis and/or thrombosis after angioplasty, carotid endarterectomy, anastomosis of vascular grafts and chronic cardiovascular devices (eg, indwelling catheters or extending "extracorporeal circulating devices"). These symptoms represent various stenotic and occlusive vascular disorders thought to be initiated by the platelet activation on the vessel walls.

PAI can be used for the prevention or termination of arterial thrombus formation, in unstable angina and arterial embolism or thrombosis, as well as for the treatment or prevention of myocardial infarction (MI) and mural thrombus formation post MI. For heart-related disorders, treatment or prevention of transient chemical attacks and treatment of thrombotic strokes or stroke-in-progress are included. PAIs may also be used to prevent platelet aggregation, embolization, or consumption in extracorporeal circulations, including enhancing animal dialysis, cardiopulmonary bypasses, hemo-perfusions, and plasmapheresis.

PAI prevents platelet aggregation, embolization, or consumption associated with intravascular devices and administration results in improved use of intra-aortic balloon pumps, ventricular assist devices, and arterial catheters.

PAI will also be useful for the treatment or prevention of venous thrombosis such as in deep vein thrombosis, IVC, renal vein or portal vein thrombosis and pulmonary vein thrombosis.

Various disorders involving platelet consumption, such as thrombotic thrombocytopenia purpura, have also been treated.

In addition, PAI can be used in many non-therapeutic applications where it is desirable to inhibit platelet aggregation. E.g. improved platelet and whole blood storage can be achieved by adding sufficient amounts of the peptides where the amount will vary depending on the length of proposed storage time, storage conditions, application of the stored material, etc.

The PAI dosage may vary depending on the desired effects and the therapeutic setting. Usual dosages are between about 0.01 and 10 mg/kg, preferably between about 0.01 to 0.1 mg/kg of body weight. Administration is preferably parenteral, such as intravenously on a daily basis for up to a week or up to one or two months or more, and this varies with peptide size. If the peptides are sufficiently small

(ie less than about 8-10 amino acid residues), other routes of administration may be used, such as intranasally, sub-lingually or the like.

Injectable mixtures can be prepared in conventional forms, either as liquid solutions or suspensions, solid forms suitable for dissolution or suspension in liquid before injection, or as emulsions. Suitable excipients are e.g. water, saline, dextrose, mannitol, lactose, lecithin, albumin, monosodium glutamate, cysteine hydrochloride or the like. In addition, if desired, the injectable pharmaceutical compositions can contain smaller amounts of non-toxic auxiliary compounds such as wetting agents, pH-buffering agents, etc. If desired, absorption-promoting preparations (e.g. liposomes) can be used.

Example 1

Analysis for platelet adhes. lonsinhibltors from snake venom

A. Description of analyzes - plate analyses

Purified platelet GP IIb-IIIa receptor was prepared as described by Fitzgerald, L.A. et al., Anal. Biochem. (1985) 151:169-177. Vitronectin receptor was prepared as described by Smith, J.W., J. Biol. Chem. (1988) 263:18726-18731. After purification, the receptors were stored in 0.1# Triton X-100 at 0.1-1.0 mg/ml.

The receptors were coated on the walls of 96-well flat-bound ELISA plates (Linbro EIA-Plus microtiter plate, Flow Laboratories) after dilution 1:200 with a solution of 20 mM Tris-HCl, 150 mM NaCl, 1 mM CaCl2, pH 7.4, to reduce the Triton X-100 concentration below its critical micelle concentration and add a 100 µl aliquot to each well. The wells were incubated overnight at 4°C and then vented to dryness. Additional sites were blocked by adding bovine serum albumin (BSA) at 35 mg/ml in the above buffer for 2 h at 30°C to prevent nonspecific binding. The wells were then washed once with binding buffer (50 nM Tris-HCl, 100 mM NaCl, 2 mM CaCl2, 1 mg/ml

BSA).

The corresponding ligands (fibrinogen, von Willebrand factor or vitronectin) were labeled with medi25l or conjugated to biotin using commercially available reagents and standard protocols. Labeled ligands were added to the receptor-coated wells at a final concentration of 10 nM (100 µl/well) and incubated for 3 hours at 30°C in the presence or absence of test samples. After incubation, the wells were aspirated to dryness and bound ligand quantified.

For 225^-mer^ed ligands, protein is solubilized at 250 µl SDS. For biotinylated ligands, bound protein is detected by addition of antibiotic antibody conjugated to alkaline phosphatase followed by addition of substrate (p-nitrophenyl phosphate) and determination of the optical density of each well at 405 nm. Reduced color development or reduced<125>I content has been observed in wells incubated with experimental samples that inhibit the binding of ligand to receptor.

B. Determination of adhesion. 1onsinhibis. 1on of raw venom

68 crude, lyophilized snake venoms obtained from either Sigma Chemical Company (St. Louis, MO) or Miami Serpentarium Labs (Salt Lake City, UT) were dissolved at 1 mg/ml in buffer (50 nM Tris, 100 mM NaCl, 0, 02# azide, 2 mM CaCl2). 1 ml aliquots of the solutions were subjected to ultrafiltration through Centrocon-10 (YM membrane) microconcentrators (Ami-con, Danvers, MA). The filtrates are mixed as test samples in the receptor/ligand analysis according to paragraph A using the GP Ilb-IIIa fibrinogen system and detection of binding using biotinylated fibrinogen. The results are shown in table 1.

The activity is present in some, but not all, species of Viperinae, but is absent in all species examined from Elapidae.

Fig. 1 shows the results of different dilutions of the filtrate for four species. Even at the highest dilution, 25 µl/0.5 ml, the three active venoms show maximum inhibition.

C. Determination of the activity of peptides in an in vivo

model for thrombosis

Purified peptides were tested for their ability to prevent the formation of thrombi in canine coronary arteries in the model described by Folts (Folts, J.D. et al., Circulation (1976) 54:365. In this model, flow reductions in a narrowed coronary artery have been shown to be caused by the formation of platelet aggregates, and agents that block the binding of fibrinogen to GP IIb-IIIa have been shown to prevent these flow reductions (Coiler, B.S. et al., Blood (1986) 68:783. The peptides were dissolved in normal saline and administered into a peripheral vein as a single bolus.

D. Effects of purified snake venom peptides on cell binding to adhesive proteins

M21 melanoma cells, which express high levels of the vitronectin receptor, were metabolically labeled bed35S-methionine, and then added to 24-well tissue culture plates coated with vitronectin. An incubation period of 1 hour at 37°C was given for cell attachment, and this was followed by a wash to remove non-adherent cells. After washing, adherent cells were solubilized and the supernatants placed in a liquid scintillation counter. The fraction of cells that remained adherent was calculated by dividing the cpm in the solubilized supernatants by the cpm in the total number of cells added to each well. The effects of purified snake venom peptides and synthetic cyclic peptides on cell adhesion were determined by incubating them with M21 cells during the incubation period.

E. The specificity of adhesion. the ion inhibition

Ultrafiltrates from three species of snake venom, Sistrurus m. barbouri, Crotalus ruber ruber and Crotalus basilicus were tested in both the fibrinogen/GP Ilb-IIIa and vitronectin/vitro-nectin receptor assays from section A. The results were assessed at different dilutions. As shown in fig. 2 the venom from Sistrurus m. barbouri preferentially inhibits the binding of fibrinogen to GP Ilb-IIIa, and the venom from Crotalus ruber ruber inhibits binding in both systems approximately equally; and the venom from Crotalus basilicus preferentially inhibits vitronectin/vitronectin receptor binding.

In the purifications described in Examples 2-6 and 8-12, PAI activity was assayed using a direct inhibition of platelet aggregation. Platelet-rich plasma (PRP) was obtained from a healthy human volunteer. Aggregation was induced by the addition of 4 µM ADP to 0.5 ml PRP in an aggregometer (Chrono-log Corp.).

A table showing the results of the amino acid composition analysis of purified PAIs according to Examples 2-6 is found after Example 6, and those showing the results for Examples 8-11 are shown after Example 8.

This analysis was accomplished by hydrolyzing the peptides using 6 N HCl and analyzing the hydrolyzate using a Beckman 121 HCl analyzer equipped with a Model 126 computer system. Cysteine acid was determined according to the method of Moore, J. Biol. Chem. (1969) 230:235-237. Tryptophan was not determined.

Example 2

Purification of platelet aggregates. lonsinhlbltor (PAI) from Eristocophis macmahoni venom

A solution of 45 mg of Eristocophis macmahoni venom (Miami Serpentarium Labs, Lot No. EM23SZ) in 1.0 ml of 0.5% trifluoroacetic acid (TFA) was chilled on ice for 20 min, centrifuged at 14,000 rpm for 3 min. to remove insoluble material and applied to a 39 mm x 30 cm, C-18 Delta Pak reverse-phase HPLC column (Waters, Milford, MA) equilibrated with 5% acetonitrile containing 1% TFA. A gradient running from 5% to 15% acetonitrile over 5 min., (2^/min.) followed by a gradient from 15% to 3056 acetonitrile over 35 min., and then to 50% acetonitrile over 20 min., was run using a Waters 600E liquid chromatograph. A flow rate of 1.5 ml/min. was maintained throughout the gradient, and the column effluent was collected for 2 min. fractions into polypropylene tubes.

The eluate from the column was recorded at 220 nm/2.5 absorbance units full scale (AUFS).

Fractions were concentrated to half their original volume using their Speed-Vac concentrator (Savant) followed by lyophilization. The samples were then reconstituted in 1 mL of distilled water and aliquots (10-50 µl) were assayed for their ability to inhibit human platelet aggregation in platelet-rich plasma induced with 20 µm ADP using a whole blood aggregometer (Chrono-Log Corp., Havertown , PA).

As shown in fig. 3, activity was found in fractions eluting at 21-25% acetonitrile concentration. These fractions were then lyophilized and rerun on C-18 HPLC column using shallower acetonitrile gradient as follows: Initial conditions consisted of 8% acetonitrile followed by a gradient to 25% acetonitrile over 68 min.

(0.25$/min.), then to 60$ acetonitrile for 10 min. One-minute fractions were collected, dried, and reassayed for platelet aggregation inhibitory activity of human platelets as above.

As shown in fig. 4 eluted the activity at 24% acetonitrile. The active fractions were then subjected to analytical HPLC with detection at 220 nm and eluted as a single symmetrical bioactive component as shown in fig. 5. Amino acid analysis of the HPLC-purified material showed that the peptide contains 49 residues including 7-8 cysteines, as indicated in Table 2.

Attempts at automated Edman degradation of the carboxyamidomethylated peptide did not provide any detectable sequence. Cleavage of this material was therefore carried out by Lys-C and Asp-N endoproteinases which produced the fragment which was sequenced and is shown in fig. 6. This analysis showed a sequence of 48 recurrences. Because of. of two tryptophan residues emerge from this frequency analysis that were not determined in the amino acid composition, the intact peptide contains 51 amino acid residues. Two Glx and an Arg residue missing from the determined sequence were presumably present at the blocked amino terminus of the peptide. Since it was very likely that one of the Glx residues was a pyroglutamyl residue at the amino terminus leading to the blocked nature of the intact peptide, the inventors removed this group from the intact carboxyamidomethylated peptide with the enzyme pyroglutamyl aminopeptidase (L-pyroglutamyl peptide hydrolase, EC 3.4.11.8, Boehringer Mannheim Biochemicals, Indianapolis, IN). Protocols described by Podell and Abraham, Biochem. Biophys. Res. Commun. (1978) 81:176-185 was used. Cleavage of 100 pg of peptide with the peptidase at a substrate-to-enzyme ratio of 100:1, followed by reverse-phase HPLC purification of the mixture on a Waters analytical C-18 column yielded material suitable for automated Edman degradation. The results of this analysis and listing of the entire sequence of this peptide, which was designated "eristicophin", are shown in Fig. 6.

The entire amino acid sequence of this PAI is shown in fig. 6. This peptide has RGD in the binding region, and shows considerable homology! with echistatin.

Example 3

Purification of PAI from Sistrurus catenatus tergeminus venom

360 mg of Sistrurus c. tergeminus venom (Miami Serpentarium Labs, Lot No. ST6SZ) was dissolved in 7.0 ml of 0.5 M acetic acid and applied to a Sephadex G-50 fine column (Pharmacia, 2.5 x 100 cm) equilibrated and eluted with 0.5 M acetic acid. The column was run at a flow rate of approximately 25 ml/hr and 5 ml fractions were collected. 25 µl of each of the fractions was pooled into groups of 10 fractions (i.e.

fractions 1-10, 11-20, etc.) and lyophilized for analysis. The dried pooled fractions were redissolved in water and aliquots assayed for inhibitory activity in ADP-stimulated aggregation of human platelets. Active fractions (31-40) were pooled and lyophilized.

This material was dissolved in 2 mL of 0.5% TFA and applied to a 19 mm x 30 cm C-18 Delta Pak reverse phase HPLC column (Waters) equilibrated with 8% acetonitrile containing 0.1% TFA. A gradient from 8$ to 30$ acetonitrile concentration over 30 min. and then to 60% acetonitrile over 20 min was run at a flow rate of 18 mL/min. The eluate from the column was collected in propylene tubes for 0.2 min. fractions and recorded at 220 nm/2.2 AUFS. The fractions were concentrated on a Speed-Vac concentrator (Savant), lyophilized and analyzed for antiaggregation activity with human platelets as previously described.

Fig. 7 shows that the PAI-containing fraction elutes at 24-25% acetonitrile. Analysis of these active fractions using HPLC with detection at 220 nm showed a symmetrical bioactive component, as shown in fig. 8. Amino acid analysis of this material revealed a peptide of 71-72 residues, including 12 cysteines, as shown in Table 2.

A portion of the purified peptide was reduced and alkylated with iodoacetamide and purified on a C-18 reverse-phase HPLC column. N-terminal sequence analysis of this material showed the following amino acid sequence for 23 cycles of Edman degradation: Glu-Ala-Gly-Glu-Glu-Cys-Asp-Cys-Gly-Ser-Pro-Ala-Asn-Pro-Cys-Cys-Asp -Ala-Ala-Thr-Cys-Lys-Leu.

The complete amino acid sequence of this PAI designated "tergeminin" is shown in Fig. 6.

The purified peptide was tested in the receptor-based assays described in 1 ex. 1, paragraph 8. The concentrations of the pure peptide at less than 100 nM inhibited the binding of Fg and vWF to GP Ilb-IIIa and of Vn and vWF to the vitronectin receptor, as shown in Fig. 9.

Example 4

Purification of the platelet aggregation inhibitor from Sistrurus milarus barbour venom

200 mg of Sistrurus m. barbouri venom (Miami Serpentarium Labs., Lot No. SM13SZ) was dissolved in 7.0 ml of 0.5 M acetic acid and applied to a Sephadex G-50 fine column (Pharmacia, 2.5 x 100 cm) equilibrated and eluted with 0.5 M acetic acid. The column was run at a flow rate of 26 mL/hr, and 5 mL fractions were collected and assayed for antiplatelet aggregation activity as previously described. Active fractions (41-50) were pooled and lyophilized. This material was redissolved in 2.0 ml of 0.5% TFA and applied to the preparative C-18 HPLC column as in Example 3 and eluted using the same gradient conditions. 2/10 minute fractions from the column were collected in polypropylene tubes, concentrated, lyophilized and assayed for platelet aggregation inhibitory activity.

Fig. 10 shows the activity profile from this HPLC column. The active fractions were subjected to analytical HPLC, which showed several fractions (45-47) that were more than 90% homogeneous. The peptide in fraction 46 (150 pg) was purified to homogeneity on an analytical C-18 column with manual collection of the symmetrical peak, as shown in Fig. 11. Amino acid analysis of this material revealed a peptide of 71-72 amino acids, including 12 cysteine residues, as shown in Table 2.

The purified peptide (150 µg) was dissolved in 300 µl reaction buffer (6 M guanidine HCl, 0.25 M Tris-HCl, 20 mM

EDTA, 20 mM dithiothreitol (DTT), pH 7.5) for 1.5 h at room temperature to reduce the peptide. This was followed by reaction with 3 µl of 4-vinylpyridine (Aldrich) at room temperature for a further hour. The reaction was stopped by the addition of 200 µl 1$ TFA and applied to an analytical C-18 HPLC column and eluted with an acetonitrile gradient in water containing 0.1$ TFA, starting at 8$ acetonitrile and running to 25$ acetonitrile for 20 min ., then to 60$ acetonitrile for 10 min.

A portion of this pyridylethylated material was subjected to N-terminal sequence analysis, as described above, and complete proteolytic cleavage of the reduced and alkylated peptide was performed using endoproteinase-Lys-C and endoproteinase Asp-N with peptide fragments isolated on either the C- 3 or C-18 reverse-phase HPLC column using acetonitrile/water/TFA gradient elution. The amino acid sequence of the N-terminus of the intact peptide and isolated proteolytic fragments was determined as described by Yarden, Y. et al., Nature (1986) 323:226, using automated Edman degradation on a gas-phase sequencer.

The entire amino acid sequence of this isolated peptide designated "barbourin" is shown in Fig. 6, together with the sequences for the proteolytic fragments. A comparison of this sequence with those from other snake venom adhesion inhibitors is shown in Fig. 12.

Example 12

Purification of PAI from Lachesis mutas venom

99 mg of Lachesis mutas venom (Miami Serpentarium Labs., Lot No. LM15FZ) was dissolved in 1 2.0 ml of 0.5% trifluoroacetic acid and cooled on ice for 20 min., centrifuged at 14,000 rpm for 3 min.

to remove insoluble material and applied to a 3.9 mm x 30 cm, C-18 Delta Pak reversed-phase HPLC column (Waters) equilibrated with 5% acetonitrile containing 0.1% trifluoroacetic acid. A gradient from 5$ to 15$ acetonitrile over 5 min. and then at 30$ over 35 min. (2$/min.) and continued to 60$ acetonitrile over 20 min. Hie drove. The flow rate was maintained at 1.5 ml/min and the column eluate was recorded at 220 nm/3 AUFS. 2 minute fractions were collected, concentrated by Speed-Vac and lyophilized. The fractions were analyzed for platelet aggregation inhibitory activity.

Fig. 13 shows the active fractions that eluted at 18% acetonitrile. These fractions were run again, on the C-18 column using a shallower gradient consisting of a 40 min. gradient from 5-28$ acetonitrile. One-minute fractions were collected, concentrated, lyophilized and assayed for platelet aggregation inhibitory activity, with the results shown in Figs. 14. These active fractions were run on an analytical C-18 column and the eluted mid-peak fraction was collected manually. The eluted material which is in a single symmetrical peak, as shown in fig. 15, was subjected to amino acid analysis and revealed a peptide of 72-73 amino acids containing 12 cysteines, as shown in Table 2.

The entire amino acid sequence of this PAI, designated "lachesin", is shown in Fig. 6.

Example 6

Purification of PAI from Crotalus viridis viridis venom

47 mg of Crotalus viridis viridis venom (Sigma Chemical Co., Lot No. 24F-0534) was dissolved in 1 1 ml of 0.5% trifluoroacetic acid, cooled on ice for 20 min., centrifuged at 14,000 rpm for 3 min.

to remove insoluble material, and applied to a 3.9 mm x 30 cm C-18 Delta Pak reverse-phase HPLC column (Waters) equilibrated with 5% acetonitrile containing 0.1% trifluoroacetic acid. A gradient from 5$ to 15$ acetonitrile over 5 min. (2$/min. ) followed by a gradient from 15$ to 30$ acetonitrile for 35 min. and then to 60% acetonitrile for 60 min. was driven. A flow rate of 1.5 ml/min. was maintained throughout the gradient, and the column eluate was collected in polypropylene tubes in two-minute fractions. The column eluate was recorded at 220 nm/3.0 AUFS. Fractions were concentrated, lyophilized and assayed for platelet aggregation inhibitory activity.

Active fractions, shown in fig. 6 as 18-19$ acetonitrile, was run on a C-18 HPLC column using a gradient of 8$-20$ acetonitrile over 48 min. ($0.25/min). Fractions were concentrated and lyophilized and tested for activity; active fractions were run on a C-18 column using 8-16% acetonitrile over 10 min., 16-20% acetonitrile over 15 min. and then to 60$ over 10 min. The eluate was recorded at 220 nm with individual peaks collected by hand in polypropylene tubes. Reanalysis of the active peak on analytical HPLC gave the results shown in fig. 17. Amino acid analysis performed on this peak revealed a 72-73 residue peptide containing 12 cysteines as shown in Table 2. The complete amino acid sequence of this PAI designated "viridin" is shown in Fig. 6 and is compared with the second PAI in fig. 12.

Example 7

Comparison of purified PAI with echistatin

The peptides purified as described in examples 2 and 4, eristicophin and barbourin, were compared with the 49-residue peptide echistatin with inhibition of fibrinogen binding to GP IIb-Illa, as described in example 1, paragraph A. Fig. 18 shows that these purified PAIs are 2 -3 times more potent in this assay than standard echistatin.

Peptides purified to homogeneity from Echis carinatus, Sistrurus m. barbouri and Eristicophis macmahoni venoms were compared with echistatin in the ADP-stimulated platelet aggregation assay. Increasing concentrations of purified snake venom peptides were added (without preincubation) at the indicated concentrations (Fig. 19). The snake venom peptides from Eristicophis macmahoni and Sistrurus m. barbouri were at least twofold more potent than echistatin, consistent with their order of potency observed for inhibiting fibrinogen binding to GP Ilb-IIIa as noted above.

Example 8

Purification of PAI from Crotalus cerastes cerastes — venom

1 g Crotalus c. cerastes venom (Miami Serpentarium Labs, Lot No. CE4SZ) was dissolved in 7.0 ml acetic acid and applied to a Sephadex G-50 fine column (Pharmacia, 2.5 x 100 cm) equilibrated and eluted with 0.5 M acetic acid. The column was run at a flow rate of 25 ml/hr with 5 ml fractions collected in polypropylene tubes. Aliquots of these fractions were assayed for aggregation inhibitory activity as previously described. Active fractions (71-80) were pooled and lyophilized. The dried material was resuspended in 2.0 ml of 0.5% TFA, insoluble material was removed by centrifugation and applied to the preparative C-18 Waters HPLC column as described in Example 3 and eluted using the gradient elution conditions described in Example 3. Fractions from the column were pooled in polypropylene tubes, concentrated and assayed for platelet aggregation inhibitory activity. Fig. 20 shows the activity profile from this HPLC fractionation.

Active fractions with platelet aggregation inhibitory activity were pooled and lyophilized and rerun on the preparative C-18 HPLC column eluted with the same gradient. Factions were merged per hand in polypropylene tubes and again analyzed for platelet aggregation inhibitory activity as before. Active fractions were analyzed on an analytical C-18 column using the conditions described in Example 4, and homogeneous fractions were pooled and lyophilized. Analytical HPLC analysis of this material is shown in fig. 21.

Purified peptide was subjected to amino acid analysis which showed that it was a peptide of 73-74 amino acids containing 12 cysteine residues, as shown in Table 3.

Purified peptide (450 µg) was dissolved in 750 µl of reaction buffer (6 M guanidine-HCl, 0.25 M Tris-HCl), 20 mM EDTA, 20 mM dithiothreitol (DTT), pH 7.50) for 1.5 h at room temperature to completely reduce the peptide followed by reaction at room temperature for 1 h with excess iodoacetamide (Fluka, 16 mg). The reaction was stopped by the addition of 500 µl of 1% TFA and applied to an analytical C-18 HPLC column and eluted with a gradient of acetonitrile from 8% to 25% acetonitrile in 20 min, then to 60% acetonitrile in 10 min. The UV-absorbing peak was collected by hand in 1.5 ml eppendorf-rlr and dried.

A portion of this carboxyamidomethylated peptide was subjected to N-terminal sequence analysis. Complete proteolytic cleavage of carboxyamidomethylated peptide was performed using endoproteinase Lys-C and the endoproteinase Asp-N peptide fragments from these cleavages were isolated on either C-3 or C-18 reverse-phase HPLC columns using acetonitrile/water/ TFA gradient elution conditions. The amino acid sequence was determined as described in example 4. The entire amino acid sequence determined for "cerastine" is shown in fig. 6, and is compared with that of other PAIs in fig. 12.

Example 9

Purification of PAI from Crotalus ruber ruber venom

1 g of Crotalus ruber ruber venom (Miami Serpentarium Labs, Lot No. CF17SZ) was dissolved in 18 mL of 0.5 M acetic acid and applied to a Sephadex G-50 fine column (Pharmacia, 2.5 x 100 cm) equilibrated at room temperature and eluted with 0.5 acetic acid. The column was run at a flow rate of 25 ml/hr with 5 ml fractions collected in polypropylene tubes. Aliquots of fractions were assayed for platelet aggregation inhibitor activity as described. Active fractions (61-70) were pooled and lyophilized. The dried material was resuspended in 2.0 ml of 0.5% TFA. Insoluble material was removed by centrifugation and applied to a preparative C-18 Water HPLC column as described and eluted using the gradient conditions described in Example 3. The fractions collected in polypropylene tubes were concentrated on a Speed-Vac concentrator and assayed for platelet aggregation inhibitory activity. .

Fig. 22 shows the activity profile after this HPLC fractionation. Individual active fractions were lyophilized. Fractions 49 and 50 were pooled and applied to the analytical C-18 reverse phase column and eluted using the conditions described in Example 4 which consisted of an acetonitrile gradient run from 8% acetonitrile to 25% in 20 min. followed for 10 min. to 69$ acetonitrile to provide homogeneous peptide which we have named "ruberin". Automated Edman degradation of carboxyamidomethylated peptide gives the sequence gene shown in fig. 6.

Example 10

Purification of PAI from Crotalus atrox

1 g of Crotalus atrox venom (Miami Serpentarium Labs, Lot No. CX16AZ) was dissolved in 10 ml of 0.5 M acetic acid and applied to a Sephadex G-50 fine column (Pharmacia, 2.5 x 110 cm) equilibrated and run at room temperature with 0.5 M acetic acid. The column was run at a flow rate of 25 ml/hour with 5 ml fractions collected in the polypropylene tube. Aliquots of the fractions were assayed for platelet aggregation inhibitory activity as previously described. Active fractions (81-100) were pooled and lyophilized. The dried material was dissolved in 2.0 ml of 0.5% TFA and applied to the preparative C-18 HPLC column and run as described in Example 3. The fractions from the column were collected in polypropylene tubes, concentrated on a Speed- Vac concentrator and assayed for platelet aggregation inhibitory activity as before. Active fractions were re-run on the analytical C-18 column to provide homogeneous peptide (Fig. 23). Amino acid analysis of this material showed that the peptide contains 72 amino acids including 12 cysteine residues, as shown in Table 3. The amino acid sequence of the isolated peptide, crotatroxin, is shown in Fig. 6.

Example 11

Purification of PAI from Bothrops cotiara

680 mg of Bothrops cotiara venom (Miami Serpentarium Labs, Lot No. B05SZ) was dissolved in 1 10 ml of 0.5 M acetic acid and applied to a Sephadex G-50 fine column (Pharmacia, 2.5 x 110 cm) equilibrated and eluted with 0.5 M acetic acid. The column was run at a flow rate of 25 ml/hr with 5 ml fractions collected in polypropylene tubes. Aliquot the fractions

were assayed for platelet aggregation inhibitory activity as previously described. Active fractions (71-90) were pooled and lyophilized. Dried material was resuspended in 2.0 mL of 0.5% TFA and applied to the Waters preparative C-18 reverse phase column. The column was eluted using the conditions described in Example 3. The fractions were collected in polypropylene tubes, concentrated on a Speed-Vac concentrator and assayed for platelet aggregation inhibitory activity. Active fractions were individually lyophilized. Several top fractions were run again on the analytical C-18 column as described in Example 4. The analytical HPLC profile of the homogeneous peptide is shown in fig. 4. Amino acid analysis of this material shows that this peptide contains 72 amino acids including 12 cysteine residues, as shown in Table 3. The complete amino acid sequence of this peptide designated "cotiarin" is shown in Figures 6 and 12.

Purified peptide was tested in the receptor assay described in example 1. The original determinations showed that low concentrations of cotiarin (1-4 nM) selectively inhibit vitronectin binding to the vironectin receptor, while the same concentrations had significantly lower inhibitory activity in the binding of fibrinogen to GP Ilb-IIIa as shown in Fig. 25, but subsequent experiments could not verify this result.

Example 12

Purification of PAI from Crotalus viridis lutosus

A. lg Crotalus viridis lutosus venom (Miami Serpentarium Labs, Lot No. CL18SZ) was dissolved in 8 ml of 0.5 M acetic acid and applied to a Sephadex C-50 fine column (Pharmacia, 2.5 x 110 cm) as was equilibrated and eluted with 0.5 M acetic acid. The column was run at a flow rate of 25 ml/hr, and 5 ml fractions were collected in polypropylene tubes. Aliquots of the fractions were analyzed for platelet aggregation inhibitory activity. The fractions (71-100) were pooled and lyophilized. Dried material was resuspended in 2.0 ml of 0.5% TFA. Insoluble material was removed by centrifugation and applied to a preparative C-18 Waters reverse-phase column and eluted using the gradient elution conditions described in Example 3. The fractions from the column were collected in polypropylene tubes, concentrated on a Speed-Vac concentrator and assayed for platelet aggregation inhibitory activity . Active fractions were lyophilized in their individual tubes. Fractions with peak activity were run again on the Waters analytical C-18 column using the acetonitrile gradient described in Example 4. The fractions were collected per hand into 1.5 ml Eppendorf tubes. homogeneous fractions were pooled and lyophilized. Analytical HPLC of this material showed a single symmetrical peak. The complete amino acid sequence of this peptide designated "lutosin" is shown in Figures 6 and 12.

B. In a similar manner to that described in paragraph A, PAI from B. jararacussu, C. basilicus, C. durissus durissus, Cv. oreganus, C. h. horridus, C v. helleri, C. durissus totonactus and from C. m. molossus isolated and purified. The amino acid compositions of several of these peptides are shown in table 3. The amino acid sequences of PAI from Ch. horridus, C. basilicus, Cm. molossus, Cv. oreganus and C. d. durissus, designated horridin, basilicin, molossin, oreganin and durissin, are shown in fig. 6. Receptor binding data for purified peptides according to examples 1-12 are shown in fig. 26.

In Examples 13-16 below, the peptides were synthesized by solid-phase techniques on an Applied Biosystems 431A Peptide Synthesizer using t-Boc amino acids activated as HOBt active esters according to instructions from the vendor, which are briefly as follows for the preparation of Boc-AAl AA(n-l)-AA(n)-O-PAM polystyrene resin.

Half mmol of selected Boc-AA(n)-O-PAM polystyrene resin is treated according to the following scheme for incorporation of Boc-AA(n-l)-OH: 1) TFA cleavage: 30% TFA in DCD, 3 min., 50 $ TFA in DCM, 16 min. 2) Washes and neutralizations: DCM washes (5X), 3 min., 5$ DIEA in DCM, 2 min., 5$ DIEA in NMP, 2 min. NMP wash (6X), 5 min. 3) Coupling: 4 equiv of Boc-AA-HOBt ester in NMP (preactivated 55 min), 38 min, DMSO to form 15$ DMS0/85$ NMP, 16 min, 3.8 equiv. DIEA, 5 min.

4) Wash and resin sample: NMP wash, 3 min.

5) Capping: 10$ acetic anhydride, 5$ DIEA in NMP, 8 min.

6) DCM washes (6X) 4 min.

Example 13

Preparation of analog No. 1 TE28L 41C64 lbarbburin ( 28- 73 ) : E- C- A- D- G- L- C- C- D- Q- C- R- F- L- K- K- G- T - V- C- R- V- A- K- G- D- W- N- D- D-T- C- T- G- Q- S- C- D- C- P- R- N- G- L - Y-G

Half mmol PAM-Gly resin (0.6 meq/g, Applied Biosystems, Foster City, CA) was subjected to method A with the necessary amino acids (introduced in order). Boe-protected amino acids had the following side chain protection: Arg(Tos), Asp(OcHex), Cys(4-MeBzl), Glu(OcHex), Lys(Cl-Z), Thr(OBzl), Trp(CHO) and Tyr (Br-Z). After generation of the complete protected peptide-resin chain, the amino-terminal Boc group was removed with TFA and the resin dried as this TFA salt form. The resin (1.3 g) was subjected to "low-high" HF stripping protocols followed by removal of HF "in vacuo". The dried peptide-resin mixture was transferred to a fritted funnel (rough) with ethyl ether and washed several times with alternating washes with ether and chloroform to remove most of the organic protecting groups and agents used in the cleavage.

The peptide mixture was transferred to 2 L of 0.4% acetic acid and pH adjusted to 7.99 with concentrated NH 4 OH. The resin was filtered from this solution and the solution was left at 4°C without stirring for 20 hours. This was followed by warming the solution to room temperature and storing it for another three days without stirring. The precipitated material was removed by filtration, and the supernatant pH-adjusted to 3.0 with acetic acid and lyophilized.

The raw material was dissolved in 8.0 ml of 0.5 M acetic acid and applied to a Sephadex G-50 fine column (2.5 x 100 cm) equilibrated with 0.5 M acetic acid. The column was run at 20 ml/h and the fractions (4 ml) were collected in polypropylene tubes. Aliquots of the fractions were dried, resuspended in water and tested for platelet aggregation inhibitory activity as previously described. Active fractions (71-90) were pooled and lyophilized.

Dried material (66 mg) was redissolved in 2.0 ml of 0.1 M acetic acid and applied to a Waters preparative C-18 column equilibrated with 8% acetonitrile containing 0.1% TFA. A gradient from 8$ acetonitrile to 20$ in 10 min. followed by a slow gradient to 30% acetonitrile for 40 min. were performed. The column was eluted at 18 ml/min and the fractions (12 sec) were collected in polypropylene tubes. The fractions were concentrated on a Speed-Vac concentrator to 1.0 ml volume, and 10 µl aliquots were examined in the platelet aggregation assay.

Active fractions (29-32) were individually lyophilized and analyzed on the analytical C-18 HPLC column with an 8-30% acetonitrile gradient. Fractions 29 and 30 were pooled and applied to the analytical column in 1.0 mL of 0.5% TFA. The main peak was collected manually and lyophilized to provide 1.6 mg of pure peptide.

Amino acid analysis of this material confirmed the identity of the peptide. Analysis of this material for its ability to inhibit the binding of fibrinogen to GP Ilb-IIIa and vitronectin to VnR is shown in Figures 26 and 27. These data demonstrate the high affinity of this analog for GP Ilb-IIIa and the relative lack of affinity at VnR for the concentrations up to 1 jjM.

Example 14

Preparation of analog No. 2.rK^^l eristicophin ( 4- 51): E- E-Y- C- T- G- K- S- C- D- C- P- R- N- P- W- N- G

Half mmol PAM-Gly resin (0.6 meq/g, Applied Biosystems, Foster City, CA) was subjected to method A with the necessary amino acids (introduced in order). Boc amino acids had the following side chain protection: Arg(Tos), Asp(OcHex), Cys(4-MeBzl), Glu(O-cHex), Lys(Cl-Z), Ser(OBzl), Thr(OBzl), Trp(CHO) and Tyr (Br-Z). Cleavage, refolding and purification of this peptide was identical to the previous examples. Receptor binding data for this analog are shown in Figures 26 and 28.

Example 15

Preparation of analog No. 3: G- C- G- K- G- D- W- P- C- A- NH2

Half mmol pMBHA resin (0.72 meq/g, Applied Biosystems, Foster City, CA) was subjected to procedure A with the necessary amino acids (introduced in order). Boc-protected amino acids had the following side chain protection: Asp(0-cHex), Cys(4-MeBzl) and Lys(Cl-Z). After completion of the preparation of the protected peptide resin, the amino-terminal Boc group was removed with TFA, and the resin was dried as the TFA salt form. The resin (1.54 g) was treated with anhydrous hydrogen fluoride (HF) containing 10% anisole, 2% ethyl methyl sulfide for 30 min. at -10°C and a further 30 min. at 0°C. HF was removed in vacuo and the peptide/resin mixture was suspended in diethyl ether followed by alternating washes with chloroform and ether 3X. After a final ether wash, the peptide was extracted from the resin with 2.0 M acetic acid, diluted with distilled water and lyophilized.

The crude peptide (370 mg) was dissolved in deoxygenated 10 mM NH4OAC, pH 8, to 0.5 mg/ml and oxidized by the dropwise addition of a small excess of 0.01 M potassium ferricyanide (K3Fe(CN)(, )-solvent- ning, stirred for an additional 20 min, and pH adjusted with acetic acid.The peptide solution was treated with Dowex AG3x4 anion exchange resin for 15 min with stirring, and the resin filtered, diluted with H 2 O, and lyophilized to provide the crude cyclized peptide .The crude cyclized peptide (392 mg) was purified by desalting on Sephadex G-25F using 0.5 M acetic acid as eluent, followed by ion-exchange chromatography on CM-Sepharose (Pharmacia) using an elution gradient formed by addn. of 100 mM NH 4 OAC to a solution of 10 mM NH 4 0 Ac, pH 4.5. Fractions having a minimum purity of 90% by HPLC analysis were pooled and lyophilized from H 2 O several times to provide 175 mg. The final purification consisted of preparative HPLC purification on a Water C-18 ref rs phase column with an acetonitrile/water/TFA gradient to provide the purified peptide. Receptor binding data for this analog are shown in Figures 26, 29 and 30.

Example 16

Preparation of further analogues

The following analogs were prepared in most cases in a manner similar to that described in Example 15. Analog 60, shown below, was prepared in solution via guanidation of the side chain of the lysine residue of analog #19 using the method of Bajusz, S. et al., FEBS Letts

(1980) 110:85-87. 1 mg of analog No. 19 was reacted with 1 mg of 1-amidino-3,5-dimethylpyrazole nitrate (Aldrich) in 1 ml of absolute ethanol in the presence of diisopropylethylamine (DIEA) at room temperature for 4 days. Product analog 60 was purified from excess reagent and starting materials by reverse-phase HPLC on a C-18 column using a gradient of acetonitrile in 0.1% trifluoroacetic acid. 900 pg of this material was isolated in purified form.

Mr

No. * 4 G-C-K-G-D-W-P-C-A-NH2

5 C-G-K-G-D-W-P-C-NH2

6 G-C-G-K-G-D-W-C-A-NH2

7 G-C-K-G-D-W-C-A-NH2

8 Acetyl-C-K-G-D-C-NH2

9 Mpr-K-G-D-Pen-NH2

10 C-K-G-D-W-P-C-NH2

11 Acetyl-C-R-G-D-Pen-NH2

No

* 12 C-K-G-D-Y-P-C-NH2

13 C-K-G-D-F-P-C-NH2

19 Mpr-K-G-D-W-P-C-NH2

34 C-K-G-D-W-G-C-NH2

35 C-K-G-E-W-P-C-NH2

36 C-Orn-G-D-W-P-C-NH2

37: C-K-A-D-W-P-C-NH2

38: C-K-AT-D-W-P-C-NH2

39: C-K-G-D-W-(Sar)-C-NH 2

40: C-K(Formyl)-G-D-W-P-C-NH241: C-K-G-D-I-P-C-NH2

42: C-K-G-D-(4-Cl-Phe)-P-NH2 43: C-K-(Sar)-D-W-P-C-NH2

44: C-K-G-D-(4-NO2-Phe)-P-C-NH245: C-K-G-D-(NMePhe)-P-C-NH246: C-H-G-D-W-P-C-NH2

47: Acetyl-C-K-G-D-W-P-C-NH248: Mpr-K-G-D-W(Formyl)-P-C-NH249: Mvl-K-G-D-W-P-C-NH2

50: Mpr-K-G-D-WT-P-Pen-NH2

51: Mpr-K-G-D-W-P-Pen-NH2

52: Mpr-K-G-D-W-P-Penr-NH2

5543: :MMpprr--KK--GG--DD-rW-W--PPT--PP<eenn-->NNHH<2>2

55: Mpr-K-G-D-W-(Thz)-C-NH 2

56: Mpr-K-G-D-H(2,4-DNP)-P-C-NH257: Mpr-K-G-D-(2-Nal)-P-Pen-NH258: Mvl-K-G-D-W-P-Pen-NH2

59: Mpr-K-G-D-W-(Pip)-Pen-NH260: Mpr-(Har)-G-D-W-P-C-NH2

61: Mpr-K-G-D-W-P-C1-NH2

62: Mpr-(D-Lys)-G-D-W-P-Pen-NH263: Mpr-(Har)-G-D-W-P-Pen-NH264: Mpr-(Acetimidyl-Lys)-G-D-W-P-C-65: Mpr-(Acetimidyl-Lys)-G-D-W-P- Pe 6 6: Mpr-(N<G>,N<G>'-ethylene-Har)-G-D-W 67: Mpr-(N<G>,N<G>'-ethylene-Har)-G-D-W

No- 68: Mpr-Har-Sar-D-W-P-C-NH2

6 9: Mpr-(Acetimidy1-Lys)-G-D-W-P-Pen-NH

70: Mpr-(Phenylimidyl-Lys)-G-D-W-P-C-NH2

71: Mpr-Har-Sar-D-W-P-PenNH2<2>

72: Mpr-(Phenylimidyl-Lys)-G-D-W-p-PenNH

73: Mpr-Har-G-D-W-(3,4-dehydro P)-C-NH2

Example 17

PAI activity of the peptides

When analyzed in the standard aggregation inhibition assay described above, analogs Nos. 3-5 had IC5Q values of 5 µM for the ability to inhibit ADP-induced human platelet aggregation. Analog No. 6, on the other hand, has an IC5Q of greater than 200 pM and analog No. 7, 100 µM. The ICsg values for the analogs of the invention in this assay are as follows:

Example 18

The activity of linear versus cyclic peptides

When tested for inhibition of fibrogen binding to GP Ilb-IIIa in the plate assay, linear RGDW-NH2 was very similar in activity to cyclic GCGRGDWPCA-NH2 (Fig. 29). In contrast, linear KGDW-NH2 was much less potent than cyclic GCGKGDWPCA-NH2 (Fig. 29). For the KGDW compounds, but not the RGDW compounds, cyclization resulted in a marked increase in the ability of the peptide to inhibit the binding of fibrinogen to GP IIb-IIIa.

Example 19

Results of plate mixing assays for synthetic peptides

The peptides prepared in Example 17, in addition to being assessed for their ability to inhibit platelet aggregation directly, were also tested in the platelet assays of the invention as described above. The results for analogs 4-8 are shown in fig. 30. As indicated in the figure, these analogs have different abilities, to varying degrees, in inhibiting the binding of fibrinogen to GP IIb-IIIa compared to vitronectin to the vitronectin receptor. Analog No. 4 appears to have the highest differential among this group. Analogues #7 and #5 on the other hand are also highly specific and have excellent platelet aggregation inhibitory activities.

Example 20

Effects of purified peptides on cell adhesion

M21 melanoma cells were labeled with<35>S-methionine and then added to vitronectin-coated plates in the presence of the indicated concentrations of the purified snake venom peptides. Cell attachment was measured by dissolving the cells remaining after an incubation and washing, as described in section C, on page 40. As shown in fig. 31 neither barbourin nor peptide 1 (abbreviated barbourin) had a significant effect on cell adhesion to vitronectin, despite both being patent inhibitors of platelet aggregation as shown in examples 2 and 3. In contrast, cotiarin which is a potent inhibitor of vitronectin binding to the vitronectin receptor very potent when it comes to inhibiting cell attachment to vitronectin. In a similar experiment, peptide #3, peptide #3 with K replaced by R (CGCRGDWPCA-NH2) and RGDS were examined on M21 cell attachment to vitronectin. As shown in fig. 32 RGDS and CGCRGDWPCA-NH2 are potent inhibitors of cell attachment while CGCRGDWPCA-NH2 was not effective up to 60 µM.

Example 21

Comparison of the analogues 60 and 10

The analogues 60 and 19 described above are peptides according to the invention containing the sequence K<*>GDX and are identical with the exception of the inclusion of K<*>. Analogue 60 has the formula: Mpr-(Har)-G-D-W-P-C-NH2;

analogue 19 has the formula:

Mpr-K-G-D-W-P-C-NH2.

These analogs were tested by standard platelet aggregation inhibition assays and using the cell adhesion assay of Example 20 above. The results are shown in figures 33 and 34. As shown in fig. 33, analogue no. 60 is effective at very low concentrations in inhibiting platelet aggregation, and is relatively less effective in inhibiting cell adhesion to vitronectin. Fig. 34 shows that analog no. 19 has good platelet aggregation inhibition activity as well as specificity, but is less active in the platelet aggregation inhibition assay than analog no. 60 counterpart. Analog No. 60 has an IC50 for platelet aggregation of approximately 0.15 nM; analogue no.

19 has an IC5Q of approximately 1 nM.

Example 22

Folf's model for canine coronary artery thrombosis

A. Initiation of cyclic flow reductions. loners (CFRs) in the open chest dog. An occluder placed on the left anterior descending (LAD) coronary artery of a 20 kg dog was performed as previously described. Phasal and mean blood flows measured by an electromagnetic (EM) flow probe and Doppler flow probe are shown in Fig. 35.

B. Effect of cvkllsk CGCRGDWPCA-NH2 (analog no. 3) on CFR in open-chest dogs. A dose of 10 mg of this peptide was infused into the peripheral vein of the dog. The blood flow patterns in the LAD as described above are shown in fig. 36. The partial ablation of the CFR as shown in the reduced slope of the flow reductions should be noted. It is also worth noting that the flow is not reduced to the same extent as in control (A). C. A second infusion of 40 mg of analog #3 was given into a peripheral vein. As shown in fig. 37, the complete ablation of the CFR indicates that complete flow has been restored in the LAD.

Example 23

Construction of expression vectors for barbourin peptides

A gene encoding full-length [L<41>] barbourine peptide (1-73) was constructed from synthetic oligonucleotides as shown in Fig. 38, which was kinased, fused and ligated into ECoRI-HindIII cleaved M13mpl8 using standard procedures. The bacterial alkaline phosphatase gene (phoA) signal sequence (Watson, M.E., Nucleic Acids Research (1984) 12:5145) was added to the barbourin construct by ligation of synthetic oligonucleotides into the EcoRI/Ncol sites of [L<41>] barbourin (1 -73) construction as shown in fig. 39. The nucleotide sequences of all constructs were verified by the Sanger dideoxy chain termination method.

A shortened version of this peptide was also constructed from synthetic oligonucleotides encoding only amino acids 28-73 of the full-length molecule. Two changes, Q<28>toE2<8>and A<64>to C^<4>, were introduced using site-directed mutagenesis as described by Kunkel et al., Meth. Enzymol. (1987) 154:367. The phoA signal sequence was added to the shortened version as described above (Fig. 40). In addition, the signal sequence for E. coli heat-stable enterotoxin II (Picken, R.W. et al., Infect. Immun. (1983) 42:269) was added to the shortened version using synthetic oligonucleotides with EcoRI and NcoI compatible ends. All bacterial secretion constructs were sub-cloned into the bacterial expression vector pPROK-1 (Brosius, J., Gene (1984) 27:151, ibid:161), available commercially from CLONTECH Lab., Inc., using EcoRI and HindIII restriction endonucleases.

A gene encoding tamdem repeats of the desired title peptide was prepared using the polymerase chain reaction (PCR) to produce the multimerization unit from full-length barbourin peptide 1-73 containing L41 and C64.

Fig. 41 shows the oligonucleotides used for the PCR synthesis. The PCR reaction was performed according to the method of Saiki, R.K. et al., Science (1988) 239:487. The resulting polymer boundary contains methionines at either end of the sequence as shown in FIG. 42 and provides desirable restriction seats for the construction.

Tandem repeats are made from the individual multimer-forming components by e.g. ligation of an EcoRI-BamHI fragment to a Bglll/Hindlll fragment in an M13mpl8 vector cleaved with EcoRI/Eindlll to form a dimer. The resulting dimer is digested with EcoRI BamHI and ligated to a BglII/HindIII fragment to produce a trimer, and so on until the desired size is achieved. This construction is shown in the diagram in fig. 43.

The multimer was then ligated into the E. coil vector pKK233-2, Amann, E. et al., Gene (1985) 40:183, available from Clontech, by digesting the vector with NcoI/HindIII and ligating a monomeric subfragment of NcoI/ BamHI and multimer subfragments of Bglll/Eindlll.

For expression as a fusion protein, the above cleaved vector was used together with an NcoI/EcoRI subfragment containing a slightly modified amino-terminal part (amino acids 1-'72) of the chloramphenicol acetyltransferase gene (Chang, C.N. et al., Gene (1987) 55:189 ) and EcoRI/HingIII subfragments of multimer constructs.

Example 24

Expression of recombinant genes

Protein expression from all of the recombinant plasmids described above is introduced according to Kanamari et al., Gene

(1988) 66:295 after transfection into appropriate E. coli host strains. The products are characterized by sodium dodecyl sulfate polyacrylamide gel electrophoresis and by their ability to inhibit ADP-induced platelet aggregation in platelet-rich plasma. After purification, multimeric proteins are converted to monomeric units by cyanogen bromide cleavage and the products analyzed as above.

Claims (13)

1. Analogous method for the production of a therapeutically active specific platelet aggregation inhibitor (PAI) peptide with the ability to inhibit the binding of Fg or vWF, GP Ilb-IIIa with substantially more, that is to say that either percentage inhibition is at least two times higher at a given concentration of PAI or that the concentration of PAI causing 50$ inhibition is at least twofold less for the Fg or vWF/GP Ilb-IIIa binding inhibition than for alternative ligand/receptor binding, i.e. potency than inhibition of the binding of vitronectin to the vitronectin receptor or fibronectin to the fibronectin receptor , where PAI has the formula: wherein K<*>is a lysyl residue of the formula: R<i>gNfCHg)4CHNHC0- , wherein each R<1>is independently H, alkyl (1-6C) or optionally one R<1>is R<2->C =NR<3>, wherein R<2> is H, alkyl (1-6C) or is a substituted or unsubstituted phenyl or benzyl residue, or NR<4>2where each R<4> is independently H or alkyl (1-6C ), and R<3>is H, alkyl (1-6C), phenyl or benzyl, orR<2->C=NR<3>is a residue selected from the group consisting of: where m is a number of 2-3, and each R<5> is independently H or alkyl (1-6C); and wherein one or two (CH 2 ) may be replaced by 0 or S provided that said 0 or S is not adjacent to another heteroatom; AAi is a small, neutral (polar or nonpolar) amino acid, and ni is a number from 0-3; AA2 is a neutral, non-polar, large (aromatic or non-aromatic) or a polar aromatic amino acid, and n2 is a number of 0-3; AA3 is a proline residue or a modified proline residue and n3 is a number of 0-1; AA4 is a neutral, small amino acid or the N-acylated form thereof, and n4 is a number of 0-3; each of X 1 and X 2 is independently a residue that can form a bond between X 1 and X 2 to achieve a cyclic compound as shown; each of and Y 2 is independently a non-interfering substituent, or may be absent; where one or more peptide bonds can optionally be replaced by a bond selected from the group consisting of —CH2NH—, -CH2S-, CH2CH2-, -CH=CH- (cis and trans), -C0CH2-, -CH(OH)CH2- and -CH2SO-; provided that if n3 is 0; is either: 1) the sum of n2 and n4 at least 2; or 2) K<*>must be different from Har or K; or 3) X2 must be different from Cys (C), penicillamine (Pen), or 2-amino-3,3-cyclopentanemethylene-3-mercaptopropionic acid (APmp); or 4) one or more peptide bonds are replaced with said alternating bond, characterized in that it comprises: starting the synthesis from the carboxy-terminal end of said peptide using an amino acid that has an alpha-amino protecting group, in which said amino acid is attached to a resin support, completing the formation of a protected peptide-resin chain, removing said protecting group, cleaving said peptide from said resin, cyclizing said peptide and optionally purifying said peptide. 2. Analogy method according to claim 1, where is H, acyl or a peptide residue or derivatized form thereof or is absent and/or Y 2 is OH, NH 2 or a peptide residue or a derivatized form thereof or is absent; and or wherein X 1 and X 2 are selected from the group consisting of cysteine (C), mercaptopropionyl (Mpr) and penicillamine (Pen); and/or wherein AA^ is G and n^ is 0 or 1; and or wherein AA2 is selected from the group consisting of the group W, F, L, Y and V; and or wherein AA2 is W; and/or wherein K<*> is K, Har, acetimidyl-Lys or phenylimidyl-Lys; and or wherein the compound of formula 1 is selected from the group consisting of or cyclic forms thereof, characterized in that the same substituted starting material is used. 3. Analogous method according to claim 2, wherein the platelet aggregation inhibitor is Mpr-K-G-D-W-P-C-NH2? characterized by using the same substituted starting material. 4. Analogous method according to claim 2, wherein the platelet aggregation inhibitor is Mpr-K-G-D-W(formyl)-P-C-NH2, characterized in that the same substituted starting material is used. 5. Analogous method according to claim 2, wherein the platelet aggregation inhibitor is Mvl-K-G-D-W-P-C-NHg > characterized in that the same substituted starting material is used. 6. Analogous method according to claim 2, wherein the platelet aggregation inhibitor is Mpr-K-G-D-W-P-Pen-NH2, characterized in that the same substituted starting material is used. 7. Analogous method according to claim 2, wherein the platelet aggregation inhibitor is Mpr-(Har )-G-D-W-P-C-NH2' characterized in that the same substituted starting material is used. 8. Analogous method according to claim 2, wherein the platelet aggregation inhibitor is Mpr-K-G-D-W-(Pip)-Pen-NH2" characterized in that the same substituted starting material is used. 9. Analogous method according to claim 2, wherein the platelet aggregation inhibitor is Mpr-(Har )-G-D-W-P-Pen-NH2, characterized in that the same substituted starting material is used. 10. Analogous method according to claim 2, wherein the platelet aggregation inhibitor is Mpr-(acetimidyl-Lys)-G-D-W-P-C-NH2, characterized in that the same substituted starting material is used. 11. Analogous method according to claim 2, wherein the platelet aggregation inhibitor is Mpr-(acetimidyl-Lys)-G-D-W-P-Pen-NH2" characterized by using the same substituted starting material. 12. Analogous method according to claim 2, wherein the platelet aggregation inhibitor is Mpr-(phenylimidyl-Lys)-G-D-W-P-C-NH2' characterized by using the same substituted starting material. 13. Analogous method according to claim 2, wherein the platelet aggregation inhibitor is Mpr-Har-G-D-W-(3,4-dehydro-Pro)-C-NH2, characterized in that the same substituted starting material is used.